How to Tell One Trial, One Measurement and One Specimen Apart in a PSLE Science Investigation

Wait, What? Three Readings Do Not Automatically Mean Three Trials

A learner measures the temperature of the same cup of water at 0 minutes, 5 minutes and 10 minutes. There are three numbers in the table, so the learner writes, “The experiment was repeated three times.”

It was not. One investigation run can produce several measurements. A single plant can also provide several measurements, and several leaves from one plant are not automatically several independent plants. If these roles are mixed up, it becomes difficult to judge what repeated evidence actually tells you.

PSLE Science investigations become much clearer when you keep three things separate: the trial, the measurement and the specimen or set-up being tested.

Quick Answer

A trial is one execution of the planned investigation under a stated condition. A measurement is one recorded observation or numerical reading made during or after that trial. A specimen is the physical object, organism, sample or set-up being investigated. One trial may contain many measurements. One specimen may be measured many times. A repeated trial may reuse a reset set-up or use a fresh comparable specimen, depending on the design.

The learner chain is: IDENTIFY THE SCIENTIFIC QUESTION → IDENTIFY WHAT COUNTS AS ONE TRIAL → IDENTIFY THE SPECIMEN OR SET-UP → IDENTIFY EACH MEASUREMENT → ASK WHAT WAS ACTUALLY REPEATED → DECIDE WHAT THE REPEATED EVIDENCE CAN SUPPORT.

Owned PSLE Science Learning Job

This guide owns one distinct job: keeping trial, measurement and specimen roles separate so repeated evidence in a PSLE Science investigation is interpreted correctly.

It does not replace the existing guides on whether an investigation needs repeated trials or more similar specimens, whether repeated measurements over time are needed, or how to organise raw observations into a results table. Those pages answer neighbouring questions. This page first makes sure the learner knows what unit is being repeated.

Why This Matters in the Current PSLE Science Frame

The 2026 PSLE Science examination assesses the 2023 Primary Science syllabus. SEAB includes scientific inquiry such as interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning. Those jobs require the learner to understand how evidence was produced.

When a table contains several rows, the rows may represent time points, conditions, trials, specimens or separate set-ups. You cannot decide what the evidence means until you know which scientific unit each row represents.

The Three Roles

Role	What it means	Example
Trial	One execution of the planned test under a stated condition	Run the cooling test once from the agreed starting temperature
Measurement	One observation or recorded value	Temperature at 0, 5 or 10 minutes
Specimen / set-up	The physical object, organism, sample or apparatus being investigated	One cup of water, one leaf, one plant, one material strip

These roles can overlap in everyday speech, which is why scientific reading must be more precise. A worksheet may not always use the word specimen; it may say object, sample, plant, animal, material, set-up or group. Ask what physical thing the measurement belongs to.

Worked Example 1: Three Time Points, One Trial

A cup of hot water is left on a table. Its temperature is recorded at 0 minutes, 5 minutes and 10 minutes.

Specimen/set-up: the cup of water.
Trial: one cooling run.
Measurements: three temperature readings.

The three readings show how the same set-up changed during one run. They do not by themselves tell you whether a second run under the same conditions would give a similar pattern.

If the question asks about change over time, the repeated measurements are useful because they reveal the time pattern. If the question asks about repeatability of the whole investigation, one run is still only one run.

Worked Example 2: Three Readings of the Same Quantity at Nearly the Same Time

A learner places a ruler beside one shoot and reads its height three times without changing the shoot or repeating the growth period.

Those may be three readings of one specimen at one stage. They can help reveal reading inconsistency, but they are not three independent growth trials.

The important question is: what was reset and run again? If nothing was reset and the biological growth period happened only once, the investigation was not repeated three times.

Worked Example 3: Three Plants, One Condition

Three similar plants are placed under the same light condition and their increase in height is measured after one week.

Now the evidence comes from three biological specimens. Depending on the investigation design, those plants may provide repeated evidence for that condition. The important distinction is that natural differences between plants are now represented in the evidence.

Measuring one plant three times does not capture the same kind of variation as measuring three comparable plants once each. Both can be useful, but they answer different evidence problems.

Worked Example 4: Several Leaves From One Plant

A learner measures the length of four leaves from one plant and says, “I tested four plants.”

The statement is false. There is one plant and four leaves. If the scientific question concerns variation among leaves on that plant, four leaves may be appropriate evidence. If the question concerns variation among plants, leaves from only one plant cannot automatically stand in for several plants.

This is an age-appropriate way to learn a deeper scientific rule: the physical source of the observations matters.

Worked Example 5: A Fresh Specimen for Each Trial

A material strip is bent until it reaches a stated condition. Bending permanently changes the strip. To repeat the test fairly, the next trial uses a fresh comparable strip.

Here each trial uses a different specimen because the first one cannot simply be reset to its original state. Reusing the damaged strip could make later trials start from a different condition.

Repeated trials do not therefore require reusing the exact same physical object. They require repeating the scientific procedure under comparable intended conditions.

Worked Example 6: The Same Apparatus Can Be Reused After Resetting

A simple set-up measures how long an object takes to travel a fixed distance under one condition. After the first run, the object is returned to the same starting position and the test is run again.

Now the apparatus and object may be reused because the intended starting state can be restored. Each run can count as a new trial if the procedure is genuinely repeated.

The scientific question is not “Did I touch the same apparatus again?” It is “Did I perform the planned test again from a comparable starting condition?”

One Trial Can Produce a Whole Curve

A graph with ten points may still come from one trial if the same set-up was measured at ten time points. More plotted points do not automatically mean more repeated experiments.

This matters when judging evidence. Ten measurements can describe one run in great detail, but they may not show how much the entire run would vary if repeated.

One Trial Can Also Produce Several Different Measurements

Suppose an investigation records both temperature and mass at the end. That is one trial with two measured outcomes, not two trials merely because two columns contain numbers.

Keep the measured quantities separate. Each outcome may support a different conclusion, and one should not be treated as a repeat of the other.

The Reset Test: A Useful Way to Recognise a New Trial

Ask: Was the planned test started again from the required starting condition?

If yes, you may have a new trial.
If the same run simply continued and another reading was taken, you have another measurement within that trial.
If another object was added but the test was not repeated, you may have another specimen inside the same design rather than another time-sequential trial.

This is a reasoning aid, not an official examination definition. Always use the design stated in the question.

Why Repetition Exists

Repeated evidence can help reveal whether a result is stable or whether measurements vary. But “repeat the experiment” should not become a memorised improvement phrase that is pasted into every investigation.

First diagnose the evidence problem. If the issue is one uncertain instrument reading, repeated readings may help. If natural differences among organisms matter, more comparable specimens may help. If the whole procedure may vary from run to run, repeated trials may help. If the tested range is too narrow, repeating the same point again may not solve that problem.

Repeated Measurement Is Not Automatically Independent Evidence

If you read the same ruler three times without moving anything, those readings share the same specimen, set-up and physical state. They may reveal reading consistency, but they are tightly connected to one another.

At Primary level, you do not need advanced statistical language to use this idea. Simply ask: What genuinely changed or restarted between the readings?

A Table Can Hide the Difference

Consider this simplified table:

Row	Temperature
1	40°C
2	35°C
3	30°C

What do rows 1–3 represent? Three time points? Three trials? Three cups? Three conditions? The numbers alone do not tell you. Read the headings, labels and surrounding method.

This is why understanding what one row represents must come before averaging, comparing or claiming repetition.

Worked Example 7: Three Trials, Three Measurements Each

An investigation is run three separate times. In every run, temperature is measured at 0, 5 and 10 minutes.

Now the structure is:

3 trials;
3 time-point measurements per trial;
9 recorded temperature values in total.

Do not call this “nine trials”. The nine numbers sit inside a two-level structure: measurements within trials.

Worked Example 8: Several Specimens Across Two Conditions

Condition A uses three similar seedlings. Condition B uses another three similar seedlings. Each seedling is measured at the start and end.

The learner must keep at least four roles visible: condition, specimen identity, start/end measurement and comparison. If all six final values are thrown into one list, the experimental structure disappears.

A good table should make it possible to tell which measurement belongs to which specimen and condition.

Earliest Weak-Link Diagnosis

Failure signature	Earliest weak link	Repair
“There are five numbers, so there were five trials.”	Evidence structure	Identify what each row/time point represents.
“I measured one plant three times, so I tested three plants.”	Specimen identity	Label the physical source of every reading.
“Three leaves from one plant are three independent plants.”	Sampling unit	Trace each observation back to its organism.
“I repeated the measurement, so I repeated the whole experiment.”	Trial definition	Ask whether the planned test was reset and run again.
“Repeating always fixes the method.”	Evidence diagnosis	Name the specific weakness repetition is supposed to address.
“More data points always mean stronger evidence.”	Evidence quality	Ask what was actually varied, repeated and sampled.

Misconception Repair: More Numbers Are Not Automatically More Independent Tests

Scientific evidence becomes useful because of how observations are produced, not simply because the page contains many numbers.

A detailed time series from one trial can reveal a pattern over time. Several specimens can reveal natural variation among specimens. Several repeated trials can reveal run-to-run consistency. These are different strengths.

Do not collapse them into one vague idea called “more reliable because there is more data”.

Question-Reading Protocol

What is the scientific question?
What physical object or organism is being tested?
What counts as one run of the planned procedure?
How many measurements are taken inside each run?
Are measurements repeated on the same specimen or taken from different specimens?
Was the set-up reset between runs?
What kind of variation can the design actually reveal?

Practice Sequence

Round 1: Label trial, measurement and specimen in simple written methods.
Round 2: Read small tables and decide what each row represents.
Round 3: Compare “three readings from one specimen” with “one reading from three specimens”.
Round 4: Compare repeated measurements within a trial with genuinely repeated trials.
Round 5: Decide which kind of repetition would address a stated evidence weakness.

Unfamiliar Transfer Challenge

A fictional material is tested under Condition X. One strip is measured every minute for five minutes. A second design uses five fresh strips, each measured once after five minutes.

Both designs produce five numbers. But the evidence structures differ. The first follows one specimen through time. The second compares several specimens at one time point.

Ask what scientific question each design can answer better. The answer depends on whether you need the time pathway or variation among specimens.

Delayed Independent Return Test

Several days later, give a new investigation containing two conditions, three specimens per condition and four time-point measurements per specimen. Ask the learner to state the number of conditions, specimens and measurements without confusing them with one another.

Then ask what would count as repeating the whole procedure. If the learner can reconstruct the evidence structure without a template, the distinction is becoming usable.

Answer-Checking Receipt

What exactly is one trial here?
What is the specimen or set-up?
How many measurements occur inside one trial?
Are several measurements from the same specimen being mistaken for several specimens?
Was the investigation genuinely reset and run again?
What variation can the repeated evidence reveal?
Does my conclusion match the evidence structure?

Common Traps

Counting data points as trials.
Counting leaves, repeated readings or time points as separate organisms without checking the design.
Calling repeated readings a repeated experiment.
Assuming a fresh specimen always means a different condition.
Repeating the wrong part of a weak method.
Averaging values before identifying what the values represent.

Parent and Tutor Teaching Guide

When a learner says “we repeated it three times”, ask, “What exactly did you repeat?” This single question often reveals whether the child means the whole investigation, one measurement or several specimens.

Draw a simple nested map: condition → trial → specimen/set-up → measurements. Do not insist on this exact nesting for every real experiment; use it only when it matches the design. The aim is to make the evidence structure visible.

Use original everyday examples rather than memorised exam phrases. Ask whether taking a photograph of the same plant three times creates three plants. Ask whether measuring one cup at three times creates three cooling experiments. These contrasts make the distinction concrete.

Useful Internal Routes

Authoritative References and Evidence Boundary

The terms trial, measurement and specimen can be implemented differently in more advanced experimental designs. This article uses age-appropriate distinctions to help Primary 5/6 learners understand how observations are organised. It is not an official SEAB marking template and does not introduce advanced statistical requirements.

The Quiet Return

Good investigation reading is not about counting how many numbers appear on the page.

Ask where each number came from. Was it another moment in the same run, another run of the procedure, or another physical specimen? Once those roles are clear, repeated evidence begins to mean something rather than merely look plentiful.

Continue learning: Return to PSLE Science: investigations, variables and fair tests, or choose English, Mathematics and Science.

Small Group Tutorials