Repeat the investigation when another trial can test whether the result is dependable, not merely because “three times” is a ritual. Keep every original reading, use the same method and conditions, compare the spread of the results, investigate anomalies and state what the evidence can reasonably support.
In Primary 5 Science tuition in Punggol, a result that matches a prediction is encouraging but not automatically reliable. One trial may reflect the expected relationship, or it may have been influenced by timing, measurement, sample variation, apparatus position or chance. Repeats help a child distinguish agreement with an idea from confidence in the evidence.
A useful Primary 5 Science tutor should explain what each repeat is checking, record all trials and connect variation to the method. Parents should see a prediction made before the test, a full results table, a decision about anomalous data and a fresh investigation where the child decides whether repetition is useful.
Choose the route that matches your question
Repeat to test dependability, not to decorate a table.Read the results
Compare trial spread and patterns.Handle anomalies
Investigate before excluding anything.Improve the method
Keep conditions and measurements consistent.Help me judge
Ask what changed after the repeats.
Open the complete chapter index
- Separate prediction from evidence
- Name the job of the repeat
- Keep the method genuinely the same
- Record every trial before judging it
- Compare spread before calculating an average
- Use an average only when it represents the trials
- Investigate anomalies without automatic deletion
- Distinguish repeat measurements from replicate samples
- Increase precision before increasing trial count
- Know when repetition will not fix bias
- Use repeats to improve the conclusion
- Stop repeating when the next trial adds little
- Design a safe home repeat investigation
- Judge tuition by the evidence trail
- Twelve-task practice route
- Parent FAQs
1. Separate prediction from evidence
A prediction states what the child expects under stated conditions. The result records what happened. Agreement is not proof that the method was reliable or the explanation complete.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
The child predicts that warmer water dissolves a sugar cube faster. One 60°C trial takes 34 seconds and one 25°C trial takes 78 seconds. The direction matches, but timing and cube size still need control.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Write the prediction before testing, then place it beside—not inside—the results table. Ask which observation supports it and which possible method error remains.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The child can accept a matching result while still identifying what another trial would test.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
2. Name the job of the repeat
Repeats may reveal random variation, confirm an observation, test measurement consistency or supply values for a representative summary. The reason should determine how trials are run and reviewed.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
Three cooling-time readings differ by one or two seconds. The repeats show a stable measurement range. Repeating with a different cup would test a different question because the apparatus has changed.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Complete the sentence “We repeat this trial to check whether…”. If the sentence names a new variable, redesign it as a separate investigation.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The child can distinguish a repeat of the same condition from a new comparison.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
3. Keep the method genuinely the same
A repeat is comparable only when relevant conditions, procedure and measurement remain consistent. Quiet changes can make the table look larger without strengthening the evidence.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
The second seedling trial receives water at a different time and is measured from the table rather than the soil line. These are not clean repeats.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Use a repeat checklist for amount, timing, starting point, apparatus position and measurement rule. Note any unavoidable change.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The child can explain why the trials are comparable and identify a condition that would invalidate the comparison.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
| Feature | Keep consistent | Reason |
|---|---|---|
| Starting quantity | Same mass or volume | Fair comparison |
| Timing | Same start and stop rule | Comparable duration |
| Measurement | Same instrument and reference | Avoid shifted readings |
| Environment | Relevant conditions controlled | Reduce alternative causes |
4. Record every trial before judging it
Deleting an inconvenient reading because it spoils the pattern reverses the scientific order. Preserve the original result, then investigate why it differs.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
Times of 42 s, 43 s and 61 s include a high value. The child keeps 61 s and checks whether the stirrer paused or the stopwatch started late.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Write results in ink or retain the original digital record. Circle an unusual value, add a method note and repeat the condition when safe and useful.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The record shows the full evidence trail and any exclusion is justified rather than hidden.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
5. Compare spread before calculating an average
An average can summarise repeated values, but it can also conceal large disagreement. Children should first inspect how close or far apart the trials are.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
Trials of 20, 21 and 22 seconds cluster closely; 10, 21 and 32 seconds share the same average but show a very different measurement story.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Order the values, find the range informally or numerically as appropriate, and discuss consistency before using a mean.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The child recognises that equal averages do not guarantee equally dependable sets.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
| Set | Values / s | Average / s | Spread |
|---|---|---|---|
| A | 20, 21, 22 | 21 | Narrow |
| B | 10, 21, 32 | 21 | Wide |
6. Use an average only when it represents the trials
A mean is useful when repeated quantitative readings measure the same condition and the data are reasonably interpretable together. It is not a repair for a flawed method.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
Five mass readings from the same balance setting can be summarised after checking them. Averaging results from different cup sizes would combine different conditions.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Label each condition, keep trial values visible, calculate only after checking comparability, and state the unit.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The child can trace the average back to valid trial values and explain why one set should not be combined.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
7. Investigate anomalies without automatic deletion
An anomalous result may arise from recording, equipment, procedure or genuine variation. The response is to inspect the evidence, not to make the table tidy.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
One plant height decreases between days. The child checks whether a different ruler zero was used, the stem bent or the value was copied incorrectly.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Review notes, apparatus and repeats; identify a plausible cause only when evidence supports it. Report uncertainty if the cause remains unknown.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The conclusion acknowledges the anomaly and calibrates confidence instead of pretending it did not occur.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
8. Distinguish repeat measurements from replicate samples
Measuring the same object several times checks measurement consistency. Testing several similar samples also captures biological or material variation. These answer related but different reliability questions.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
Measuring one leaf three times checks the ruler method. Measuring five leaves under the same condition examines variation between leaves.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Label the table “same sample measured again” or “new sample under same condition”. Choose the design that fits the claim.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The child knows what source of variation the repeats can and cannot reveal.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
9. Increase precision before increasing trial count
Many repetitions of a poorly defined procedure may reproduce confusion. Sometimes the best next move is clearer timing, a more suitable scale or a fixed measuring point.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
Ten cooling readings taken at “about five minutes” are weaker than three readings taken exactly five minutes after heating with the thermometer at the same depth.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Identify the largest uncontrolled measurement decision, standardise it, then repeat.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
Variation falls for a reason connected to the improved method, not simply because more numbers were collected.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
10. Know when repetition will not fix bias
Repeats reduce uncertainty from variable trials but do not correct a systematic error that affects every reading in the same direction.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
A balance reads 5 g too high each time. Repeating produces consistent but biased masses.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Use a known reference or zero check, compare instruments when appropriate and separate consistency from accuracy.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The child can explain how results can agree closely and still be wrong.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
| Problem | Would repetition reveal it? | Better check |
|---|---|---|
| Random timing variation | Often | More controlled trials |
| Consistent zero error | Not necessarily | Calibrate or zero instrument |
| Changed sample size | Creates new variation | Standardise samples |
| Transcription mistake | May look anomalous | Check original record |
11. Use repeats to improve the conclusion
The conclusion should reflect both the direction of the results and their consistency. Strong language requires stronger evidence.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
All four trials show faster dissolving in warmer water, but times vary moderately. The child states that the tested warmer condition consistently had a shorter time, not that temperature always determines an exact time.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Name the tested conditions, summarise the repeated pattern, mention important variation and avoid universal claims.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The claim is supported by all retained results and stays within the investigation design.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
12. Stop repeating when the next trial adds little
More trials use time and materials. A sensible stopping decision considers safety, available time, variation and the purpose of the task.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
After four closely clustered readings with a clear method, a fifth identical trial may add less learning than testing whether the method transfers to a new condition.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Before starting, set a planned number or stopping rule appropriate to the classroom task. Change it only with a reason.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The child can justify why the evidence is sufficient for the limited conclusion without claiming certainty.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
13. Design a safe home repeat investigation
Home practice should use safe materials and small quantities. The learning target is recording and comparison, not creating dramatic effects.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
Compare the time for equal ice cubes to melt on two safe surfaces, repeating each condition with cubes of similar size while avoiding electrical equipment and slippery floors.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Have an adult supervise, define the timing rule, photograph starting sizes if helpful and record all trials. Stop if the setup is unsafe or uncontrolled.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
The child produces a transparent table and can explain one limit of the design.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
14. Judge tuition by the evidence trail
A polished table is not enough. Parents need to see the original prediction, first result, reason for repeating, all trials, method notes and changed conclusion.
This matters in Primary 5 Science because the visible answer is only the last part of the learning. The learner also has to notice the relevant information, choose a route, carry it out and check whether the result still fits the task. When parents focus on those decisions, the original concern—why an experiment is repeated even when the first result matches the prediction—becomes something that can be observed and improved rather than guessed about.
A concrete example
The tutor shows that a first matching result prompted two repeats, one anomaly led to a timing-rule repair, and an unseen task required the child to choose between calibration and repetition.
The example should be kept as evidence. Ask the child to explain what was noticed first, what was decided next and where help entered. A correct final response after a large prompt and a correct response produced independently are different pieces of evidence. Both can be useful, but they should never be recorded as though they show the same level of control.
What to do next
Ask what uncertainty the repeats addressed, what remained, and what the child decided independently on the next investigation.
Keep the next task close enough for the same idea to apply, then change one feature so that memory of the previous surface cannot carry the child. If the child succeeds, change a second feature. If the child stalls, return only the smallest prompt that restores the decision. This creates a clear route from supported practice to independent science work.
What counts as progress
Progress appears in better experimental decisions, honest records and appropriately cautious conclusions.
Parents can record four simple facts: whether the child started without a hint, whether the chosen method fitted the question, whether the work stayed accurate, and whether the child could explain or check the result. Those four observations are more useful than a vague report that the lesson “went well”. They also make the next tuition conversation calm, specific and fair.
| Artifact | Useful evidence | Limit |
|---|---|---|
| Prediction | Expectation stated first | Not a result |
| All trials | Variation visible | May share one bias |
| Method notes | Explains changes | May be incomplete |
| Unseen design | Transfer of judgement | One context |
A twelve-task practice route for the next fortnight
Practice task 1: Separate prediction from evidence
Begin with one short task built from the example in this chapter: The child predicts that warmer water dissolves a sugar cube faster. One 60°C trial takes 34 seconds and one 25°C trial takes 78 seconds. The direction matches, but timing and cube size still need control. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Write the prediction before testing, then place it beside—not inside—the results table. Ask which observation supports it and which possible method error remains. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The child can accept a matching result while still identifying what another trial would test. Keep the record short enough that it can guide the next lesson.
Practice task 2: Name the job of the repeat
Begin with one short task built from the example in this chapter: Three cooling-time readings differ by one or two seconds. The repeats show a stable measurement range. Repeating with a different cup would test a different question because the apparatus has changed. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Complete the sentence “We repeat this trial to check whether…”. If the sentence names a new variable, redesign it as a separate investigation. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The child can distinguish a repeat of the same condition from a new comparison. Keep the record short enough that it can guide the next lesson.
Practice task 3: Keep the method genuinely the same
Begin with one short task built from the example in this chapter: The second seedling trial receives water at a different time and is measured from the table rather than the soil line. These are not clean repeats. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Use a repeat checklist for amount, timing, starting point, apparatus position and measurement rule. Note any unavoidable change. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The child can explain why the trials are comparable and identify a condition that would invalidate the comparison. Keep the record short enough that it can guide the next lesson.
Practice task 4: Record every trial before judging it
Begin with one short task built from the example in this chapter: Times of 42 s, 43 s and 61 s include a high value. The child keeps 61 s and checks whether the stirrer paused or the stopwatch started late. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Write results in ink or retain the original digital record. Circle an unusual value, add a method note and repeat the condition when safe and useful. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The record shows the full evidence trail and any exclusion is justified rather than hidden. Keep the record short enough that it can guide the next lesson.
Practice task 5: Compare spread before calculating an average
Begin with one short task built from the example in this chapter: Trials of 20, 21 and 22 seconds cluster closely; 10, 21 and 32 seconds share the same average but show a very different measurement story. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Order the values, find the range informally or numerically as appropriate, and discuss consistency before using a mean. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The child recognises that equal averages do not guarantee equally dependable sets. Keep the record short enough that it can guide the next lesson.
Practice task 6: Use an average only when it represents the trials
Begin with one short task built from the example in this chapter: Five mass readings from the same balance setting can be summarised after checking them. Averaging results from different cup sizes would combine different conditions. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Label each condition, keep trial values visible, calculate only after checking comparability, and state the unit. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The child can trace the average back to valid trial values and explain why one set should not be combined. Keep the record short enough that it can guide the next lesson.
Practice task 7: Investigate anomalies without automatic deletion
Begin with one short task built from the example in this chapter: One plant height decreases between days. The child checks whether a different ruler zero was used, the stem bent or the value was copied incorrectly. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Review notes, apparatus and repeats; identify a plausible cause only when evidence supports it. Report uncertainty if the cause remains unknown. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The conclusion acknowledges the anomaly and calibrates confidence instead of pretending it did not occur. Keep the record short enough that it can guide the next lesson.
Practice task 8: Distinguish repeat measurements from replicate samples
Begin with one short task built from the example in this chapter: Measuring one leaf three times checks the ruler method. Measuring five leaves under the same condition examines variation between leaves. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Label the table “same sample measured again” or “new sample under same condition”. Choose the design that fits the claim. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The child knows what source of variation the repeats can and cannot reveal. Keep the record short enough that it can guide the next lesson.
Practice task 9: Increase precision before increasing trial count
Begin with one short task built from the example in this chapter: Ten cooling readings taken at “about five minutes” are weaker than three readings taken exactly five minutes after heating with the thermometer at the same depth. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Identify the largest uncontrolled measurement decision, standardise it, then repeat. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: Variation falls for a reason connected to the improved method, not simply because more numbers were collected. Keep the record short enough that it can guide the next lesson.
Practice task 10: Know when repetition will not fix bias
Begin with one short task built from the example in this chapter: A balance reads 5 g too high each time. Repeating produces consistent but biased masses. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Use a known reference or zero check, compare instruments when appropriate and separate consistency from accuracy. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The child can explain how results can agree closely and still be wrong. Keep the record short enough that it can guide the next lesson.
Practice task 11: Use repeats to improve the conclusion
Begin with one short task built from the example in this chapter: All four trials show faster dissolving in warmer water, but times vary moderately. The child states that the tested warmer condition consistently had a shorter time, not that temperature always determines an exact time. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Name the tested conditions, summarise the repeated pattern, mention important variation and avoid universal claims. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The claim is supported by all retained results and stays within the investigation design. Keep the record short enough that it can guide the next lesson.
Practice task 12: Stop repeating when the next trial adds little
Begin with one short task built from the example in this chapter: After four closely clustered readings with a clear method, a fifth identical trial may add less learning than testing whether the method transfers to a new condition. Remove the answer and any completed working before the child begins. Give quiet thinking time, and ask the child to mark the exact point at which a decision became difficult. This keeps the practice diagnostic; it reveals the barrier instead of covering it with immediate help.
After the attempt, use this response: Before starting, set a planned number or stopping rule appropriate to the classroom task. Change it only with a reason. Then set a fresh parallel task on another page. Count the task as secure only when the child can begin, continue and check with no more support than the agreed prompt. The evidence to look for is: The child can justify why the evidence is sufficient for the limited conclusion without claiming certainty. Keep the record short enough that it can guide the next lesson.
A practical parent decision
Repeat when the extra trial can answer a real reliability question. Keep the method and conditions comparable, preserve every reading, inspect spread before averaging, investigate anomalies and recognise that repetition cannot cure a systematic bias. Stop according to a planned classroom purpose, then test whether the child can choose and justify repeats in a new investigation.
- Was the prediction written before the result?
- What uncertainty is the repeat checking?
- Were relevant conditions and measurement rules consistent?
- Are all trials visible?
- Was spread inspected before averaging?
- Does the conclusion match the repeated evidence?
The Punggol Science Article Index remains the broad hub. Read why school and tuition experiments may differ for the neighbouring question about results produced by different setups.
Parent questions answered
Must every experiment be repeated three times?
No. The number should suit the question, time, safety and variation. Three is not a universal scientific rule.
If all trials match exactly, is that always good?
It may show consistency, but also check instrument resolution, copied values and systematic error.
Should an anomalous result be crossed out?
Keep the original, investigate possible causes and justify any treatment in the analysis.
When should my child calculate an average?
When repeated quantitative readings represent the same condition and an average is meaningful for the task.
Does repetition make an experiment fair?
Not by itself. Fairness also depends on variables, controls, procedure and measurement.
Can repeats prove a prediction correct?
They can strengthen evidence for a limited conclusion under tested conditions, not prove a universal claim.
What is the difference between accuracy and consistency?
Consistent results agree with one another; accurate results are close to the appropriate true or reference value.
Can we practise safely at home?
Yes with simple, supervised comparisons using familiar materials and no heat, chemicals or electrical risks.
When should we stop repeating?
Use a planned rule based on task purpose, variation, safety and time. More trials are not automatically more educational.
What should I ask the tutor?
Ask what the repeat checked, whether the method stayed comparable, how variation changed the conclusion and what unseen design tested judgement.
Current official references and useful next reading
- Punggol Science Article Index
- Primary Science: Why School and Tuition Experiments Differ
- MOE: Primary Science syllabus
- SEAB: PSLE Science (0009) for examination from 2026
Official curriculum and examination links were checked on 8 October 2026. School sequencing can vary, so parents should compare the child’s current scheme of work and subject level before treating any example here as the next compulsory topic.
