How to Read “Repeat Until” in a PSLE Science Method Without Counting Every Loop as a New Trial

Wait, What? Repeating a step does not automatically mean repeating the experiment.

A method can say “repeat until…”, “continue adding…”, “measure every minute…” or “keep doing this until the value no longer changes”. A learner sees the word repeat and concludes that the investigation has many trials.

But the repeated action may still belong to one run. It may create several measurements inside that run. It may use one specimen throughout. It may continue only until a stopping condition is reached.

The key question is not “How many times did something happen?” The key question is:

What was reset and started again?

Quick Answer

Identify the scientific question and the starting state of one run.
Find the action that is repeated.
Find the stopping condition: what makes the loop end?
Ask whether the setup returns to a fresh starting state before the sequence begins again.
If the same run simply continues, you may have repeated steps or measurements, not new trials.
If the whole test is reset and performed again under the same intended condition, that can be a repeated trial.
If a different similar organism or object is used, that may be another specimen rather than another repeat of the same specimen.
Use the wording and structure of the actual question. Do not impose one universal counting rule on every investigation.

The Exact PSLE Science Learning Job This Guide Owns

This guide owns one job: reading repeated procedural loops correctly so the learner can distinguish steps, measurements, trials, specimens and stopping conditions.

It complements, rather than replaces, the separate guide on one trial, one measurement and one specimen. That article defines the evidence units broadly. This one focuses on the special method structure in which an action repeats inside a run until an endpoint or criterion is reached.

For the 2026 PSLE, SEAB states that Science assesses the 2023 Primary Science syllabus. The assessment objectives include applying scientific inquiry, interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning. Reading how a method actually produces evidence is therefore part of the scientific job.

Five Things That Can Repeat — and They Are Not Interchangeable

What repeats?	What it means	What it does not automatically mean
Procedure step	An action is performed again inside the same run	A new trial has begun
Measurement	The same or related quantity is recorded again	The whole investigation was repeated
Trial	The defined test sequence is restarted from an appropriate starting state	A new specimen must be used
Specimen	Another organism, object or sample is included	Every specimen experienced an independent full trial unless the method says so
Cycle or loop	A set of actions repeats until a stopping rule is reached	Each loop is independent evidence

Strong inquiry reasoning keeps these units separate because they answer different questions about the evidence.

The Reset Test

When you are unsure whether a new trial has begun, use the reset test:

Was the apparatus or specimen returned to the defined starting condition?
Was the tested condition set again?
Did the sequence that produces the result begin again?
Would the new result count as another independent run of the same comparison?

If the answer is no and the method is simply continuing from the current state, you are probably still inside the same run.

The word probably matters. Scientific methods define their own units. If a question explicitly defines a trial or repeat, follow that definition.

The Stopping Condition

“Repeat until” has two parts:

the loop: what action is repeated;
the stopping condition: the observable or measurable rule that ends the loop.

Examples of stopping conditions in original practice structures might include:

until the reading reaches a stated value;
until no further change is observed over a defined interval;
until a visible endpoint appears;
until a fixed amount has been added;
until a stated number of cycles has been completed.

A good stopping condition should be applied consistently. If two set-ups are stopped according to different hidden rules, the comparison may no longer be fair or interpretable.

Worked Example 1: Repeated Measurements Inside One Run

Imagine an original practice investigation. A container of water is allowed to cool. Its temperature is recorded every two minutes until it reaches a stated temperature.

There may be many temperature readings:

Time	Temperature
0 min	70°C
2 min	64°C
4 min	59°C
6 min	55°C

These are repeated measurements of the same changing system. The container was not returned to 70°C after every reading. The whole run did not restart. Counting four readings as four trials would misdescribe how the evidence was produced.

Worked Example 2: Repeating an Addition Until an Endpoint

Imagine a method that says: add one drop, mix, observe, and repeat until a visible endpoint appears.

Each add–mix–observe cycle is a loop within the run. The number of loops may itself become a result. But that does not mean each drop created an independent trial.

To perform a second trial, the investigator would normally need to begin again with a fresh appropriate starting state under the same intended conditions, unless the question defines the method differently.

Worked Example 3: A Repeated Trial Really Does Reset

Imagine a toy car released from the same marked position on a ramp. Its travel distance is measured. The car is returned to the marked start and the full procedure is repeated three times.

Here the reset is visible:

same starting position;
same intended test condition;
same measured outcome;
full sequence restarted.

Those are repeated trials, not merely repeated measurements within one continuous run.

Worked Example 4: Repeated Observations of One Specimen

A plant is observed once a day for six days. The learner records height and leaf appearance.

Six days of observations do not automatically mean six specimens or six independent trials. They may be a time series from one organism.

This distinction matters because repeated measurements of one specimen tell you how that specimen changes over time. Several similar specimens tell you something different about variation between individuals.

Why This Matters for Averages

Learners often see several numbers and assume they should be averaged. But an average is meaningful only when you know what the numbers represent.

Several independent trials under the same condition may sometimes be summarised.
Measurements taken at different times show a time pattern and should not be collapsed into one number if the pattern matters.
Measurements from different test conditions should not be averaged together if the purpose is to compare those conditions.
Several values produced inside one “repeat until” loop may represent the path to an endpoint rather than repeated estimates of the same quantity.

Before calculating anything, identify the evidence unit.

Why This Matters for Fair Tests

A loop can quietly change the starting state. Suppose the same object is tested again and again without being allowed to return to its original condition. Later loops may not be directly comparable with earlier ones.

Ask:

Does each loop alter the specimen?
Does material accumulate or get used up?
Does temperature, position, moisture or another relevant state carry over?
Does the stopping condition itself make one run longer than another?

If carryover matters, the method may need a reset, a fresh specimen or a different comparison design.

The Evidence-Production Map

For difficult methods, draw this on rough paper:

START STATE → ACTION → OBSERVE/MEASURE → STOP?
NO → repeat the action
YES → record the run result
RESET?
YES → next trial
NO → investigation ends or moves to another stage

This makes the loop structure visible without memorising a special vocabulary.

The PSLE Science Inquiry Reading Protocol

QUESTION: What scientific relationship is being investigated?
START: What is the defined starting state?
CHANGE: Which condition is deliberately changed, if any?
MEASURE: What observation or measurement answers the question?
LOOP: Which step or measurement repeats?
STOP: What criterion ends the loop?
RESET: What must return to the starting state before another trial?
UNIT: What counts as one trial, one measurement and one specimen in this method?
COMPARE: Which results can validly be compared?
CONCLUDE: What can the evidence support, and what can it not support?

Failure Signatures

The learner counts every repeated step as a new trial.
The learner counts every reading in a time series as an independent repeat.
The learner assumes several specimens automatically mean several repeated trials.
The learner says “repeat for accuracy” without explaining what should be repeated or reset.
The learner averages values that came from different time points or test conditions.
The learner ignores a carryover effect because the word “repeat” sounds like a fresh start.
The learner cannot state the stopping condition.
The learner treats “until no further change” as if it means the same fixed duration for every set-up.

Earliest Weak-Link Diagnosis

Sequence failure: the learner cannot identify which action is inside the loop.
Evidence-unit failure: trial, measurement and specimen are mixed.
Reset failure: the learner does not notice that the system continues from its current state.
Stopping-rule failure: the learner cannot tell when the loop ends.
Comparison failure: values from unequal evidence units are compared as if they were equivalent.
Method-evaluation failure: the learner proposes “repeat more” without knowing what evidence problem repetition is meant to solve.

How to Improve a Repeat-Until Method

When a question asks you to evaluate or improve a method, do not automatically add more repeats. First diagnose the weakness.

If the endpoint is vague, define an observable stopping criterion.
If the added amount per loop differs, standardise the increment.
If the observer decides “no change” too quickly, define the observation interval.
If the system carries over into the next intended trial, reset or use a fresh appropriate specimen.
If the result varies between independent runs, repeated trials may help reveal that variation.
If the question needs the path over time, repeated measurements may be more useful than one endpoint.

The improvement must repair the evidence problem while preserving the scientific question.

Common Misconceptions

“Repeat means trial.” Not necessarily.
“More measurements mean more reliable independent evidence.” Measurements from one run may be strongly dependent on one another.
“A trial needs a new specimen.” Some methods can reset the same object; others need fresh similar specimens. The design decides.
“Same duration is always fair.” Some methods compare at a fixed time; others stop at a common endpoint. Those are different scientific designs.
“Until no change” means the process has stopped completely. It may only mean no detectable change under the method and interval used.

Retrieval and Practice Sequence

Take three methods from your own practice materials.
Underline the start state, loop action and stopping condition in different ways.
Count measurements, trials and specimens separately.
Explain what must reset before another trial can begin.
Redraw the method as a small loop diagram.
Change one feature: replace a fixed-time endpoint with a repeat-until endpoint, or vice versa. Explain how the evidence changes.
Return several days later and repeat on an unfamiliar method without your labels.

Unfamiliar Transfer Test

Transfer the skill across very different surface contexts: a cooling setup, repeated observations of a living specimen, a mechanical movement, a light or sound investigation, or a classification procedure with repeated checks.

You understand the method structure when you can identify the loop and evidence units even when the topic changes.

Delayed Independent Return Test

After a delay, read a fresh investigation and answer six questions without notes:

What begins one run?
What repeats inside the run?
What is measured?
What stops the loop?
What must reset for a new trial?
Which results may be compared?

If you can answer all six and explain why, the distinction is becoming usable rather than memorised.

Answer and Checking Receipt

Start state: I know where one run begins.
Loop: I know what repeats.
Measurement: I know what each reading represents.
Stopping condition: I know what ends the loop.
Reset: I know what creates a genuinely new run.
Specimen: I have not confused another specimen with another measurement.
Comparison: I am comparing like evidence units.

Parent and Tutor Teaching Guide

When a child says “there are five trials because the step happened five times”, ask the learner to point to where the whole method restarts.

Use physical counters if needed. Put one large card down for a trial, smaller counters inside it for repeated measurements or loop actions, and a separate counter for each specimen. This makes the nested structure visible.

Then remove the counters and use a different scientific context. Ask the learner to reconstruct the structure verbally: “One run begins here, this step repeats, this condition stops it, this reset creates the next trial.”

Do not teach “repeat three times” as a universal improvement phrase. Ask what source of uncertainty, variation or method weakness the repetition is meant to address.

Useful Internal Routes

Authoritative References and Evidence Boundaries

The examples are original teaching examples. Terms such as “trial” and “repeat” must be interpreted from the method and question in front of the learner; this guide does not create an official PSLE counting rule or marking formula.

Quiet Return

Science evidence has a history. A number came from a measurement. A measurement happened inside a run. A run began from a starting state. A loop may have repeated many times before the run ended.

When you keep that history visible, “repeat until” stops being a confusing instruction and becomes a map of how the evidence was actually produced.

Small Group Tutorials