The first trial warms the spoon.
The second trial begins with the same spoon.
The child says, “That is fair because I used the same spoon both times.”
But the spoon is no longer in the same starting state.
Using the same object does not guarantee a fair comparison if the first trial changes what the object is like before the second trial begins.
This guide belongs to the Primary 4 Science Learning Hub. It develops a distinct investigation skill: recognising order effects and carryover, deciding whether a trial can be reset, and knowing when a fresh comparable specimen is safer than pretending the system has returned to its original state.
The teaching level remains Primary 4. The page does not require formal counterbalancing designs or advanced experimental statistics. It builds the foundational idea that a fair comparison needs comparable starting conditions.
Quick Answer: The Reset-and-Carryover Loop
TRIAL 1 → ASK WHAT CHANGED → IDENTIFY WHAT MAY CARRY OVER → DEFINE THE REQUIRED STARTING STATE → RESET IF POSSIBLE → VERIFY THE RESET → RUN TRIAL 2 → COMPARE ONLY IF STARTS ARE COMPARABLE
This is an eduKate teaching routine, not an official MOE examination formula.
Wait, What? The Order of Testing Can Become Another Variable
Suppose two materials are tested on the same cup.
Material A is tested first while the cup and water are freshly prepared.
Material B is tested second after the cup has already warmed, some water has evaporated, and the thermometer has been moved.
If B gives a different result, was the difference caused by material—or by the fact that B was tested second?
When the first trial changes the starting state of the second, test order itself becomes scientifically relevant.
1. What is carryover?
For this guide, carryover means that a change from one trial remains present when the next trial begins.
Possible carryover includes:
- heat remaining in a spoon, cup or surface;
- water remaining on an object after wetting;
- a stretched material not returning to its earlier shape;
- a plant already damaged by the first treatment;
- a screen or card moved and not returned accurately;
- residue from one material remaining before another is tested;
- a sensor or thermometer still responding to the previous condition.
The carryover may be obvious or subtle.
2. Same object can be an advantage—and a danger
Using the same object can reduce differences between specimens.
For example, measuring the same plant before and after one day avoids comparing two naturally different plants.
But the same object becomes a problem when the first test changes it irreversibly or for too long.
The scientific decision is not:
“Same object is always fair.”
It is:
“Does using the same object preserve a comparable starting state for the next trial?”
3. A reset is not just putting the apparatus back in the same place
A child returns the spoon to the same mark on the table.
Position has been reset.
Temperature may not have been reset.
A true reset restores the scientifically relevant starting conditions, not merely the visible arrangement.
Ask:
- Was temperature restored?
- Was volume restored?
- Was moisture removed?
- Was the same starting distance restored?
- Was the object physically unchanged?
- Was the sensor returned to the same reference condition?
4. Resetting a heat investigation
Trial 1 places a metal spoon in warm water.
The spoon warms.
Trial 2 is supposed to compare another water condition.
If the warm spoon is used immediately, the second trial does not begin from the same spoon temperature.
Possible repair:
- allow the spoon to return to the defined starting condition;
- check its starting temperature where appropriate;
- or use a fresh comparable spoon if the teacher-designed method permits.
The correct route depends on the question and safety.
5. Resetting a cooling comparison
Suppose one cup is tested with foam, then the same cup is quickly changed to cloth.
Problems can carry over:
- cup temperature has changed;
- water volume may have changed;
- some foam may remain around the cup;
- the second wrapping may be applied differently because the cup is already warm;
- the surrounding surface may be warmer from the first run.
A stronger design may use two comparable cups prepared at the same starting temperature and volume.
That creates a different trade-off: the cups themselves must be sufficiently comparable.
6. Resetting a shadow investigation
Light trials often appear easy to reset because the objects can simply be moved back.
But a good reset checks:
- source position;
- screen position;
- object orientation;
- distance reference points;
- room lighting;
- measurement line on the shadow.
If one object is rotated slightly between trials, the next result may reflect orientation rather than distance.
7. Resetting a volume investigation
Water is poured from Cylinder A to Container B and back again.
If droplets remain in B, the second trial begins with less water.
If the question expects the same starting volume, carryover from retained droplets matters.
The solution is not to invent the missing volume.
Use a carefully designed method that preserves or re-establishes the intended starting amount.
8. Resetting a plant investigation may be impossible
A plant has several roots deliberately damaged in a teacher-provided example.
Can the same plant be “reset” to an undamaged state for the next trial?
No.
The change is not reversible within the investigation.
This is a case where separate comparable specimens may be needed.
Never damage living things unnecessarily merely to create multiple trials. Teacher-provided data or safe observational examples may be preferable.
9. Reversible vs irreversible changes
A useful P4 distinction:
| Change | Likely reset? | Example |
|---|---|---|
| Position | Often | Move card back to marked location |
| Timer state | Usually | Reset to 0 s |
| Temperature | Sometimes, with time/control | Allow object to return to defined condition |
| Wetness | Sometimes | Dry fully if scientifically appropriate |
| Broken root | No | Damage cannot be undone |
| Spilled water | Not by pretending | Prepare a fresh starting volume |
Reset decisions depend on the property that matters to the question.
10. Carryover can create a false trend
Imagine testing three water temperatures in order:
- cool;
- warm;
- hot.
The same metal object is transferred rapidly from one to the next without returning to the same starting temperature.
The object becomes progressively warmer even before each new test begins.
A smooth trend may appear, but some of it can come from accumulated carryover.
Neat patterns still need a fair method.
11. Carryover can hide a real effect
Trial 1 leaves a cup very warm.
Trial 2 tests a material expected to reduce cooling.
Because the cup starts unusually warm, the second result may look less comparable.
Carryover can exaggerate, reduce or reverse apparent differences.
That is why the scientific problem is the changed starting state, not one particular direction of error.
12. Order effect vs natural variation
Two trials differ.
Possible reasons:
- carryover from the first trial;
- naturally different specimens;
- measurement variation;
- different apparatus;
- real effect of the tested condition.
Do not label every difference “order effect”.
Use evidence.
13. The reset checklist
Before the next trial, ask:
| Starting condition | Restored? |
|---|---|
| Temperature | □ |
| Volume / amount | □ |
| Position / distance | □ |
| Orientation | □ |
| Moisture / residue | □ |
| Object integrity | □ |
| Instrument state | □ |
| Timing reference | □ |
Use only the rows relevant to the investigation.
14. Verify the reset
Do not assume reset succeeded.
Examples:
- check the marked distance;
- confirm timer = 0;
- confirm starting volume;
- confirm thermometer starting reading if relevant;
- confirm object is dry if dryness matters;
- confirm sensor is back in its reference position.
A reset that is not checked may remain only an intention.
15. Fresh specimens can solve one problem and create another
Using a fresh specimen avoids carryover.
But two specimens may differ naturally.
Example:
Plant A and Plant B are not identical.
One may begin taller, have more leaves or have different root development.
Therefore choosing fresh specimens requires attention to comparability.
This leads directly into Same Specimen Before–After vs Different Similar Specimens.
16. When order cannot be neutral
Some investigations naturally move through time.
Example:
measure a plant on Day 1, Day 3 and Day 5.
The order is part of the question.
You should not “reset” the plant to Day 1.
The correct method tracks change over time.
The key is to distinguish:
- time sequence intentionally being studied;
- unwanted carryover that contaminates a comparison.
17. Sequential observation vs repeated trial
Sequential observation:
follow the same system through time.
Repeated trial:
recreate the same starting condition and run the comparison again.
These are different scientific jobs.
Confusing them creates unnecessary resets or false comparisons.
18. Treatment order can matter
Imagine a sponge tested dry first, then wet.
If the second test uses the already-wet sponge, the order may be part of the intended comparison.
But if the scientific question asks about two independent dry-start conditions, the wet carryover is a problem.
Ask:
“What state is each trial supposed to begin in?”
19. Do not “reset” by changing the question
If one material cannot be restored, a pupil may substitute another object without noting the change.
Now specimen identity has changed.
That may be a valid new method, but it should be stated.
A reset preserves the intended question.
A replacement can create a parallel-specimen design.
20. Instrument carryover
Not only specimens carry effects forward.
Instruments can too.
Examples:
- thermometer still warm from previous cup;
- sensor still stabilising;
- measuring cylinder contains droplets from previous liquid;
- balance not returned to zero;
- timer not reset.
Batch 21’s Instrument Checks, Zeroing and Reference Tests develops instrument-state checking.
21. Data carryover
A pupil copies the previous trial’s value into the next row before measuring, intending to “change it later”.
Now the record itself can carry over an old value.
Use a fresh row and record only after observation.
Carryover can occur in physical systems and in data handling.
22. Expectation carryover
After seeing Trial 1, the learner expects Trial 2 to be higher.
That expectation may affect a fuzzy measurement.
Use fixed definitions and independent checks where useful.
Batch 21’s Observer Expectations, Confirmation Bias and Independent Checks owns that evidence-quality job.
23. Original Carryover Casebook
Case 1 | Warm spoon
Second trial begins with pre-warmed spoon.
Problem: temperature carryover.
Repair: restore defined start or use appropriate comparable fresh object.
Case 2 | Wet card
A card is tested dry, then immediately tested again while damp.
Problem: material state changed.
Case 3 | Moved screen
Screen position changes during reset.
Problem: geometry no longer comparable.
Case 4 | Spilled liquid
Second trial uses remaining liquid.
Problem: volume carryover.
Case 5 | Root damage
Same plant expected to serve as undamaged control afterward.
Problem: irreversible change.
Case 6 | Timer not reset
Trial 2 begins at 47 s on display.
Problem: timing reference carryover.
Case 7 | Sensor moved
Sensor orientation differs after reset.
Problem: measurement-system state changed.
Case 8 | Residue
Material from Trial A remains on cup during Trial B.
Problem: treatment contamination.
Case 9 | Day-by-day plant growth
Plant is not reset daily.
Not a problem: the question intentionally studies one continuing system over time.
Case 10 | Two fresh cups
Used to avoid thermal carryover.
New check: cups must be comparable.
Case 11 | Previous result shown
Observer knows expected Trial 2 direction.
Problem: expectation can carry forward even when apparatus resets.
Case 12 | Reset verified
All relevant starting conditions checked before Trial 2.
Strength: order effect is reduced for the tested factors.
24. The Reset Decision Table
| Question | If yes | If no |
|---|---|---|
| Did Trial 1 change the specimen? | Identify carryover | Continue checking |
| Can the relevant state be restored? | Reset and verify | Consider fresh comparable specimen |
| Would a fresh specimen create large natural variation? | Record baseline / match specimens | Use parallel design if appropriate |
| Is the sequence itself the scientific question? | Do not reset the evolving system | Preserve comparable starts |
| Can the reset be checked? | Verify before next trial | Limit confidence |
25. What this guide does not require
Primary 4 learners do not need:
- counterbalancing terminology;
- Latin-square designs;
- washout-period statistics;
- formal repeated-measures models;
- advanced causal-inference notation.
The foundational habit is:
ask what the first trial changes before trusting the second.
26. Original Practice Set
- What is carryover?
- Why can the same object make a comparison unfair?
- What is a true reset?
- Why is putting apparatus back in the same place sometimes insufficient?
- Give one heat-related carryover example.
- Give one irreversible biological example.
- Why can fresh specimens solve one problem and create another?
- What is the difference between sequential observation and repeated trial?
- Why should a reset be verified?
- How can a timer create carryover?
- How can a sensor create carryover?
- Why is a day-by-day plant study not supposed to reset the plant?
- What does a spill do to the next trial?
- Why can residue matter?
- What should happen if the starting state cannot be restored?
- Why should the question remain unchanged when resetting?
- Can expectation carry over even after physical reset?
- What is one useful reset check for a shadow experiment?
- When is using a fresh specimen reasonable?
- Write one sentence describing an order-effect problem.
27. Practice Answers
1. A change from one trial remains and affects the next trial’s starting state.
2. The first trial may change the object, so the second begins from a different state.
3. Restoring all scientifically relevant starting conditions, not merely appearance.
4. Temperature, moisture, volume or instrument state may still differ.
5. A spoon remains warm from Trial 1.
6. Damaged roots cannot be restored to undamaged condition.
7. Fresh specimens can differ naturally.
8. Sequential observation follows one evolving system; repeated trial recreates a starting condition.
9. An intended reset may be incomplete.
10. If not returned to zero or started at the same event, elapsed times differ.
11. It may remain warm, oriented differently or still stabilising from the previous condition.
12. The continuing change over time is the thing being studied.
13. Volume or other starting properties may no longer match.
14. One treatment may remain present in the next condition.
15. Use a fresh comparable specimen or redesign and state the limitation.
16. Otherwise the learner may accidentally run a different investigation.
17. Yes. Knowledge of the earlier result can affect later judgement.
18. Re-check source, object and screen positions against marks.
19. When carryover cannot be safely or reliably removed and comparable specimens are available.
20. Example: “The second cooling trial began with a warmer cup because the first trial changed the cup’s starting temperature, so treatment and order are confounded.”
28. The Carryover Diagnostic
| If the learner… | Likely weak link | Repair |
|---|---|---|
| says “same object means fair” | starting-state reasoning | ask what Trial 1 changed |
| resets position only | hidden state | check temperature/moisture/volume |
| uses fresh object casually | natural variation | establish comparability |
| resets evolving time study | question type | distinguish sequence from repeat |
| assumes reset worked | verification | measure starting condition |
29. A 40-Minute Carryover Lesson
Minutes 1–5: identify what Trial 1 changes in six scenarios.
Minutes 6–10: sort reversible vs irreversible changes.
Minutes 11–15: design a reset checklist.
Minutes 16–20: inspect a same-spoon heat comparison.
Minutes 21–25: compare same-specimen vs fresh-specimen options.
Minutes 26–30: diagnose a hidden timing reset error.
Minutes 31–35: identify a case where no reset should happen.
Minutes 36–40: write a bounded fair-comparison conclusion.
30. What Parents and Tutors Can Ask
- “What did the first trial change?”
- “Is that change still present?”
- “What must be restored before the next trial?”
- “How will you verify the reset?”
- “Would a fresh specimen be fairer?”
- “What new variation would a fresh specimen introduce?”
- “Are you repeating a trial or following one system through time?”
- “Could the order itself be affecting the result?”
31. Continue Batch 23
- Same Specimen Before–After vs Different Similar Specimens
- Treatment Duration, Observation Time and Measurement Intervals
- Rate of Change, Change Per Unit Time and Fair Time Comparisons
The Quiet Return
The learner prepares Trial 2.
Before touching the apparatus, one question now comes first:
“What has Trial 1 left behind?”
If the answer is heat, moisture, damage, changed volume, movement, residue or expectation, the next trial does not yet begin from the same scientific place.
