PSLE-SCI-REALITY-0096
Wait, What? A sensor can respond correctly and the conclusion can still be too strong.
A fictional test strip turns purple when Substance Q is present. A sample turns purple, and the label on a classroom poster says: “PURPLE = Q.”
Then a second fact appears. Substance R can also make the same strip turn purple under some conditions.
The observation has not disappeared. The strip really changed colour. The problem is the inference: one visible response has been treated as if it could have only one possible cause.
Reality Lab Vol No.096 teaches a durable evidence-transfer habit: when a sensor, test or indicator responds, ask whether the response is selective enough to identify the claimed target uniquely.
Quick Answer
- Record exactly what the sensor or test did.
- Separate the response from the explanation for that response.
- Ask what else can produce a similar signal.
- Check blanks, known references and relevant competing substances or conditions.
- Look for evidence about selectivity or interference.
- Where the claim matters, compare with another appropriate measurement principle if available.
- State only what the response can uniquely support.
What This Page Owns — and What It Leaves With Specialist Owners
This is not a chemistry lesson about sensor design, a medical-testing guide or a forensic interpretation page. Its job is one Primary 5/6 transfer problem: a real-world communication object treats one sensor response as though no other cause could produce it.
- Reality Lab Vol No.006 | “It Happened on Camera” — Does a Demonstration Test Why It Happened?
- Reality Lab Vol No.049 | “The Test Found Nothing” — Did a Positive Control Show It Could Find Something?
- Reality Lab Vol No.094 | “Detected” — Does That Tell You How Much Is There?
Vol No.094 asks whether a positive signal supports an amount. Vol No.096 asks a different question: does this signal point specifically enough to the target being claimed?
Original Reality Lab Case: The Purple Patch Sensor
This case is fictional and constructed for teaching. It makes no claim about a real product, substance or health test.
A paper sensor has a pale patch. In reference tests:
| Reference | Sensor result |
|---|---|
| Blank liquid | Stays pale |
| Substance Q | Turns purple |
| Substance R | Turns purple |
| Substance S | Stays pale |
An unknown turns purple.
What is directly supported? The unknown caused a response that is consistent with Q and R under the conditions tested. What is not yet supported? “The unknown definitely contains Q and cannot be explained by R.”
Observed, Claimed and Inferred
| Layer | Statement |
|---|---|
| Observed | The sensor patch turned purple. |
| Reference evidence | Known Q and known R can both produce purple in this fictional test. |
| Claimed | The unknown contains Q. |
| Hidden inference | Q is the only plausible cause of the purple response. |
The hidden inference is the part that needs testing.
Response Is Not Identity
A sensor measures or responds to some physical or chemical feature. The public claim may then translate that response into an identity: “this means Q”. That translation is strong only if other plausible causes have been ruled out sufficiently for the intended use.
The scientific habit is to keep a gap between the two sentences until the selectivity evidence closes it.
What Is an Interference?
In measurement science, an interference is something other than the intended target that can affect detection, identification or measurement. For Primary 5/6, the useful question is simply: what else could make this test look positive?
An interference might increase a signal, reduce it, change colour, alter a sensor surface, disturb a measurement or imitate part of the target response. The details depend on the method.
The Alternative-Cause Table
| Possible explanation | What would help distinguish it? |
|---|---|
| Q caused the response | A known-Q reference and another appropriate Q-sensitive check |
| R caused the response | Reference test with R and a method that separates Q from R |
| Background or contamination caused the response | Blank controls and clean handling checks |
| Instrument or strip malfunction | Known negative and positive references |
Worked Case 1: The Light Sensor and Two Light Sources
A sensor’s output rises when Source A is switched on. But Source B at a different wavelength can also raise the output. A response therefore proves that the detector received a signal it can respond to; it does not by itself identify which source produced it unless the sensor or method can discriminate between them.
Worked Case 2: The Gas Indicator
A fictional indicator changes from blue to yellow in the presence of Gas X. Humidity can also cause a smaller colour shift. A yellow result may support “the indicator changed under this sample”, but a strong “Gas X is definitely present” claim needs evidence that humidity or other relevant conditions cannot explain the observed change.
Worked Case 3: The Sensor Is Highly Selective
Suppose many plausible competing substances are tested and do not produce the target response, while Q does. An independent method using a different physical principle agrees. Now the Q interpretation is much stronger.
The lesson is not “sensors cannot identify anything”. It is “identity claims depend on how specifically the measurement responds”.
Worked Case 4: The Blank Is Positive
If a blank sample also triggers the sensor, the problem changes immediately. The positive result may be coming from contamination, background, a reagent or the apparatus. A blank does not tell you every possible source of error, but it tests an important alternative explanation.
Worked Case 5: Two Different Methods Agree
Method A responds to one property of Q. Method B measures the same target property using a substantially different physical principle. If both measurements support the same conclusion on the same material, a method-specific interference becomes less likely. NIST describes such different-principle checks as orthogonal measurements in advanced measurement work.
What Evidence Would Strengthen a Specific Identification?
- A known positive reference produces the expected signal.
- A blank remains negative.
- Relevant possible interferences have been tested.
- The sensor response differs enough between target and non-target materials.
- The result is repeated under appropriate conditions.
- An independent measurement based on another principle supports the same interpretation.
- The claim is kept within the validated sample and operating conditions.
What Would Weaken It?
- Several different materials can produce the same response.
- The test has never been checked against plausible competing causes.
- The blank also gives a positive result.
- The indicator is affected strongly by temperature, humidity, colour or another uncontrolled condition.
- A dramatic signal is treated as unique proof without showing selectivity evidence.
- The result is interpreted outside the conditions for which the method was evaluated.
Tempting Reasoning That Fails
- “The sensor worked, therefore the target is proven.” A working sensor can still respond to more than one cause.
- “Any interference makes the method useless.” No. Scientists identify, control and account for interferences; useful methods do not need to be magical.
- “A bigger signal means a more certain identity.” Signal size and uniqueness are different questions.
- “Two tests agreeing always proves the result.” If both depend on the same interference, agreement can still be misleading. Different measurement principles can be especially valuable.
How Far Can the Conclusion Travel?
If a test is known to respond to Q and R, a careful sentence might be: “The sample produced a response consistent with Q or R under this test.” If additional selective evidence distinguishes Q, the statement can become narrower and stronger.
Science becomes more precise by ruling out plausible alternatives, not by pretending alternatives never existed.
PSLE-Style Transfer Case
A fictional indicator turns orange when Material M is present. Tests show that high temperature can also turn it orange even without M. An unknown sample tested at high temperature becomes orange.
Question: Why is the conclusion “Material M is present” not yet secure?
Reasoned answer: The orange response has at least two possible causes in the stated conditions: M or high temperature. A fairer test should control temperature or use another appropriate measurement that can distinguish M from the temperature effect.
Explained Practice
Practice A: A blank gives the same colour as the unknown. What should you suspect? Background, contamination, reagent behaviour or another method problem may be contributing to the response.
Practice B: Q and R both trigger Sensor A, but only Q triggers Sensor B. Both sensors are positive on the unknown. What became stronger? The case for Q, because the second response helps distinguish the competing explanation R.
Practice C: A product says its sensor is “specific” but gives no interference information. What question should follow? Which non-target substances or conditions were tested, and how did the sensor respond?
Delayed Independent Return: The S-I-G-N-A-L Check
- S — Signal: What actually changed?
- I — Interferences: What else can cause that response?
- G — Guides and references: Do known positives and blanks behave correctly?
- N — New method: Would another appropriate principle distinguish the alternatives?
- A — Alternative explanations: Which remain alive?
- L — Limit the claim: Say only what the evidence uniquely supports.
Parent and Tutor Teaching Guide
Use a doorbell analogy. Hearing the bell proves the bell circuit produced a sound, but not automatically who pressed it. The camera, a second witness or another clue may distinguish the cause.
Then return to science: measurements give signals, and explanations connect signals to causes. Strong reasoning asks whether more than one cause can fit the same observation.
Authoritative Sources
- Singapore Examinations and Assessment Board — 2026 PSLE Science Syllabus
- Ministry of Education Singapore — Primary Science Teaching and Learning Syllabus 2023
- National Institute of Standards and Technology — Interferences
- National Institute of Standards and Technology — Orthogonal Measurements
NIST describes interferences as non-target substances that can affect detection, identification or quantitation, and describes orthogonal measurements as measurements using different physical principles to reduce method-specific bias or interference. For a Primary 5/6 learner, the transferable habit is simple: a response is evidence, but the cause still has to earn its identity.
The Quiet Return
The sensor may be telling the truth about what it sensed.
The mistake can happen one sentence later, when we decide what that signal must mean.
When a test lights up, ask what else could light it up too.
