PSLE Science Reality Lab Vol No.118 | “The Sample Label Matches” — Can We Prove This Is Still the Same Sample?

PSLE-SCI-REALITY-0118

Wait, What? The Number Can Be Perfectly Measured From the Wrong Bottle

Imagine two clear bottles on a laboratory bench. Both contain colourless water. Both have white caps. One came from Pond A. The other came from Pond B. A technician measures Bottle A very carefully and gets a beautifully repeatable result.

There is only one problem: the bottle was accidentally swapped before analysis.

The measurement can still be technically excellent. The instrument can be calibrated. The blank can be clean. Duplicate readings can agree. The number can even be reported with sensible uncertainty. Yet the scientific conclusion about Pond A fails if the tested bottle was really Pond B.

This is why scientific evidence needs identity as well as measurement. A sample must remain connected to the place, object, time and collection event that the report says it represents.

A label helps. A chain-of-custody record helps more. Neither is magic. They are pieces of evidence about the sample’s journey.

Quick Answer

Identify the sample using a unique code, not only a description such as “pond water”.
Trace where the sample was collected, when it was collected and who handled it.
Check each handoff: transport, receipt, storage, preparation, splitting and analysis.
Look for mismatched labels, missing times, broken seals, unexplained relabelling or gaps in the record.
Separate two questions: Is this the same sample? and Did this sample remain scientifically suitable during its journey?
Keep the conclusion narrow. A complete custody record supports identity and documented handling; it does not automatically prove that the sample was chemically unchanged or that the final measurement is correct in every other respect.

The Exact Learner Job This Page Owns

This Reality Lab page owns one transfer problem: evaluating whether a scientific result is genuinely linked to the sample, place, object or event that the public claim says was tested.

It does not become the general owner of chain-of-custody systems. The wider eduKate explanation of custody, possession and handoffs remains with Chain of Custody | Who Had the Shipment, When and Under What Condition. This page applies the same evidence discipline to a Primary 5/6 scientific sample.

It also does not replace PSLE Science owners for record keeping, sampling, variables, measurement, sample storage or method limitations. Those routes remain available here:

Original Reality Lab Case: Four Bottles, One Missing Handoff

This is an original composite teaching case. It does not reproduce a real investigation, exam question or commercial laboratory record.

A pupil team collects four water samples from four fictional locations. Each bottle receives a code.

Sample code	Collection location	Collection time	Field note
R-201	North inlet	08:10	Clear water
R-202	East bank	08:25	Some sediment
R-203	South outlet	08:40	Clear water
R-204	West bank	08:55	Floating plant fragments nearby

The field sheet records that all four bottles were placed in one insulated box. At 10:30 the box reached the laboratory. Three bottles were logged correctly. The fourth bottle had a damp label. Someone rewrote the code as “R-203?” and placed it beside another bottle while checking the paperwork.

Later, a report presents one striking result and says it came from the south outlet.

Can we trust the number? That is not yet the first question.

The first question is: which physical bottle was actually analysed?

Observation, Claim and Hidden Inference

Layer	Example
Direct record	A bottle entered the instrument under code R-203.
Public claim	“The south outlet contained 42 units.”
Hidden inference	The analysed bottle was definitely the same physical sample collected at the south outlet at 08:40.
Additional hidden inference	The sample remained suitable during transport and storage.
Further hidden inference	The method measured the intended quantity well enough for the conclusion.

A custody record mainly strengthens the first hidden inference: identity and documented handling. It does not automatically settle the others.

Why Sample Identity Is a Scientific Variable Even When Nobody Calls It One

In a school fair test, pupils learn to control variables so the comparison is meaningful. In real scientific work, sample identity plays a similar protective role. If Sample A becomes Sample B by mistake, then the measured value is attached to the wrong experimental condition.

The error may be invisible in the final spreadsheet. A row can still contain a neat number, a unit and a timestamp. The spreadsheet cannot repair an identity break that happened earlier.

That is why good science preserves identity from the physical world into the data record.

What Does Chain of Custody Actually Add?

A chain-of-custody record is a chronological record of who had responsibility for a sample and when control changed. U.S. EPA guidance describes chain of custody as a way to trace a sample from collection through later handling and analysis. In scientific work, this helps make the sample’s history reconstructable.

A useful record may include:

unique sample identifier;
collection site or source;
date and time;
collector;
container type or count where relevant;
transport conditions;
who transferred the sample;
who received it;
storage location;
sample splitting or subsampling;
preparation or pretreatment;
analysis performed;
unusual conditions, damaged labels or broken seals.

The record does not need to look dramatic. Its power comes from continuity.

The Provenance Check: Can We Reconstruct the Route?

Provenance means the documented origin and history of an evidence object. For a scientific sample, ask whether you can travel backwards from the final result to the physical collection event.

Which result row are we looking at?
Which laboratory sample code produced that result?
Which container carried that code?
Where and when was that container collected?
Who handled it after collection?
Were there any relabelling, splitting or transfer steps?
Can those steps be matched by time and identity?

If a link is missing, the correct response is not automatically “the result is false”. It is “the identity claim is less well supported than it would be with a continuous record”.

A Matching Label Is Evidence, Not Proof of an Unbroken Journey

Suppose the bottle entering the instrument says R-203. That is useful evidence. But labels can be copied, rewritten, damaged or attached incorrectly. A stronger system combines labels with independent records: field sheets, receipt logs, timestamps, seals, container counts and documented transfers.

The point is not to become suspicious of every label. The point is to understand why serious scientific systems do not rely on one fragile evidence link when sample identity matters.

A Complete Custody Record Does Not Prove the Sample Was Unchanged

This is one of the most important boundaries in the article.

A custody record may show that Bottle R-203 was collected, sealed, transferred, received and stored without an identity gap. Yet the water inside it could still change chemically or biologically while it waits. That is why Vol No.106 owns the separate question of holding time and sample stability.

Identity and stability are different evidence jobs:

Question	Main evidence
Is this still the same physical sample?	Identifiers, custody records, seals, handoffs, logs
Did the contents remain suitable for the intended measurement?	Holding time, preservation, temperature, container conditions, stability evidence
Was the sample measured correctly?	Method, calibration, controls, uncertainty, QC
Does the sample represent the larger system?	Sampling design and representativeness

A strong scientific conclusion may need all four questions answered.

The Seal Check: What Does an Intact Seal Tell You?

A tamper-evident seal can strengthen confidence that a container was not opened without leaving visible evidence. It does not guarantee that the original sample was collected correctly, that the seal itself was applied to the correct bottle, or that temperature and holding-time requirements were met.

This is a general Reality Lab lesson: every evidence object has a job boundary. A seal is good at one job. It does not become evidence for every other job simply because it looks official.

The Handoff Check: Gaps Matter More Than the Number of Signatures

Imagine a form containing six signatures but no time recorded between the third and fourth transfer. Compare it with a simpler form that clearly records each transfer, time, container count and condition.

The stronger record is not necessarily the one with more handwriting. It is the one that lets a reader reconstruct what happened.

The Split-Sample Problem: One Bottle Can Become Several Evidence Objects

A laboratory may divide one sample into several portions. Perhaps one portion is tested immediately, one is stored, and one is sent to another laboratory. Once that happens, each new portion needs its own identity while retaining a documented link to the original sample.

A public report that says “the same sample was tested by two laboratories” should therefore preserve how the split occurred. If the portions were not mixed adequately before splitting, they might not contain the same distribution of particles or material even though they share the same parent sample.

The Representation Check: A Timeline Can Be More Honest Than a Badge

Scientific communication sometimes compresses provenance into a badge such as “verified sample” or “secure chain”. A more informative display is often a short timeline:

08:40 collected → 09:05 sealed in field box → 10:30 received by laboratory → 10:42 logged → 11:15 refrigerated → 14:10 subsampled → 15:00 analysed.

The timeline gives the reader something that can be checked. The badge gives only a conclusion.

Alternative Explanation 1: A Label Was Copied Incorrectly

A handwritten 3 becomes an 8. A barcode is attached to the wrong tube. A temporary code is copied into the wrong spreadsheet row. These are ordinary human-system failure possibilities, not accusations of misconduct.

Independent identity checks reduce the chance that one small mistake controls the whole evidence chain.

Alternative Explanation 2: The Right Sample Was Tested but the Wrong Site Was Entered

The physical bottle can remain correct while metadata—the information describing it—becomes wrong. A result can therefore be scientifically measured from the correct bottle but mapped to the wrong location in the final database.

This is why data provenance continues after the physical sample reaches the instrument.

Alternative Explanation 3: Nothing Was Swapped, but the Sample Changed

Identity can be perfect while stability is poor. A three-day delay, warm transport, light exposure or unsuitable container can change some quantities. Do not use a complete custody form as a substitute for holding-time evidence.

Alternative Explanation 4: The Sample Is Right but the Conclusion Travels Too Far

Even a perfectly identified sample from the south outlet does not automatically describe the whole pond. Provenance answers “where did this evidence come from?” Representativeness answers “how much of the larger system can this evidence describe?”

What Evidence Would Strengthen the Identity Claim?

Unique identifiers assigned at collection.
Field records matching identifiers, times and locations.
Documented transfers between people or controlled locations.
Receipt logs showing sample condition and container count.
Seals or other integrity controls where appropriate.
Documented relabelling rather than silent replacement of labels.
Clear records of sample splitting, subsampling and preparation.
Database entries that preserve the link between laboratory code and field code.
Investigation of discrepancies rather than simply overwriting them.

What Would Weaken It?

Two different samples share the same identifier.
A container appears with a code not present on the field sheet.
A handoff time is impossible or missing.
A sample count changes with no explanation.
A damaged label is rewritten without recording the correction.
A bottle is split but the daughter portions are not traceable to the parent sample.
The final database location differs from the field record and nobody explains why.

Worked Case 1: The Perfect Instrument, Wrong Bottle

An instrument measures a reference standard correctly, duplicates agree and the sample result is precise. Later, camera footage shows two bottles were exchanged before analysis. The measurement-quality evidence remains real, but it applies to the wrong sample identity. The public location claim must therefore be corrected.

Worked Case 2: The Label Smudged but the Chain Survives

A bottle label becomes partly unreadable, but its barcode, sealed transport bag, field sheet, container count and laboratory receipt record all agree. The damaged label does not automatically destroy the evidence chain because several independent identity records still connect the bottle to its origin.

Worked Case 3: A Complete Custody Record, Poor Storage

Every handoff is documented perfectly, but the sample spent eight hours in conditions unsuitable for the target measurement. Identity is strong. Sample suitability may not be. The correct conclusion separates those two facts instead of calling the whole process either “good” or “bad”.

Worked Case 4: Two Labs, One Parent Sample

One bottle is mixed and split into Portion X and Portion Y. X goes to Lab 1 and Y to Lab 2. Their results differ. The custody records help prove that both portions came from the same parent sample. They do not explain the disagreement. The next questions concern splitting, preparation, method and measurement performance.

Tempting Reasoning That Fails

“The label matches, so identity is certain.” A label is evidence but can be wrong or copied incorrectly.
“The custody form is complete, so the result must be accurate.” Custody mainly supports identity and documented handling, not measurement accuracy.
“A broken seal proves tampering.” It is a warning that requires investigation; damage can have more than one explanation.
“Any paperwork gap makes the sample scientifically useless.” The effect depends on what the gap prevents us from establishing and what other evidence survives.
“If the bottle is the right one, the sample still represents the place perfectly.” Representativeness is a separate sampling question.

How Far Can the Conclusion Travel?

A strong chain of identity allows a report to say with greater confidence, “This result came from this identified sample collected at this stated place and time.” It does not automatically justify, “This result describes every location,” “the sample never changed,” “the instrument was perfect,” or “the explanation is proven”.

Scientific trust grows when each part of the claim has the right kind of evidence.

PSLE-Style Transfer Case

Three plant samples are labelled A, B and C. Their leaves look very similar. During transport, label B falls off. At the laboratory, someone reattaches it to the middle container because that seems most likely. The three containers are then measured carefully.

Question: Why should the result assigned to Sample B be treated cautiously even if the measuring instrument worked correctly?

Reasoned answer: The sample identity became uncertain when the label was lost and was reattached by guess rather than by independent evidence. The measurement can be correct for the material in the container while the claim that it came from original Sample B remains uncertain.

Explained Practice

Practice A: A bottle has a unique barcode, but the field sheet lists the wrong collection time. What should happen? The mismatch should be investigated and documented rather than silently edited. The barcode supports physical identity; the time metadata still needs reconciliation.

Practice B: A custody record is perfect, but the sample exceeded its recommended holding time. Can identity evidence fix that? No. Identity and stability are different questions.

Practice C: Two sealed containers have identical printed labels. Is that a strength? No. Unique identifiers are stronger because identical labels create ambiguity about which physical sample is which.

Delayed Independent Return: The I-D-E-N-T-I-T-Y Check

I — Identifier: Is the sample uniquely labelled?
D — Documentation: Does the field record match the physical container?
E — Exchanges: Are handoffs recorded?
N — No unexplained gaps: Is any period of possession or storage unclear?
T — Transfers and transformations: Were splits, relabelling or preparation steps recorded?
I — Integrity controls: Are seals, container counts or condition records relevant and intact?
T — Tested object: Can the final result be traced back to the physical sample?
Y — Your conclusion: Does the claim stay within what identity evidence actually supports?

Parent and Tutor Teaching Guide

A useful home activity needs no laboratory equipment. Place three identical cups on a table and label them A, B and C. Ask the learner to leave the room while you swap two cups but keep the labels in place. When the learner returns, ask, “If I measure Cup A perfectly, have I proved I measured the original Cup A?”

Then repeat the activity using unique coloured seals or a written handoff record. The goal is not to teach bureaucracy. It is to show that a scientific number needs an identity trail back to the world.

For a more advanced learner, give four evidence cards—label, seal, field sheet and laboratory receipt—and remove one. Ask how much confidence changes and what additional evidence could repair the missing link.

Authoritative Sources

The current PSLE Science assessment objectives require learners to interpret and analyse information, evaluate observations, information and methods, and communicate explanations and reasoning. The Primary Science syllabus also asks learners to maintain healthy scepticism, consider uncertainty and understand how Science is communicated. Sample identity is a perfect example: the learner must ask not only “What number was measured?” but also “What physical evidence object produced that number?”

The Quiet Return

A scientific result is not only a number.

It is a number attached to an object, place, time and method.

Before asking whether the measurement is convincing, make sure you know what was actually measured.

Small Group Tutorials