PSLE-SCI-REALITY-0094
Wait, What? A test can be good enough to say “I found it” but not good enough to say exactly how much is there.
A fictional sensor checks water for Substance Q. The sensor gives a small but repeatable signal above its blank reading. A dashboard displays a green badge: Q DETECTED.
Then someone writes underneath: “The water contains 0.037 units of Q.”
That second sentence sounds like a natural extension of the first. It is not necessarily one. Scientific methods can have different levels of evidence. A signal may be distinguishable enough to support a detection decision while still being too small, noisy or weakly calibrated to support a reliable numerical amount.
Reality Lab Vol No.094 teaches one durable transfer habit: when a real-world report says something was detected, separate the evidence for presence from the evidence for quantity.
Quick Answer
- Ask what signal the method actually observed.
- Check what the blank or background normally looks like.
- Find out what evidence is required before the method calls a result “detected”.
- Then ask a different question: over what range can the method measure amount with acceptable reliability?
- Do not turn a yes/no detection into an exact concentration merely because software prints a number.
- Keep the conclusion attached to the tested sample and method.
- If amount matters, look for calibration and measurement-quality evidence that supports quantification.
What This Page Owns — and What It Leaves With Existing Science Owners
This is not a laboratory-chemistry manual and it does not re-teach the full concepts of sensitivity, uncertainty, calibration or precision. Its job is narrower: it applies PSLE Science evidence reasoning to a common real-world communication jump — from “the test detected a signal” to “therefore we know the exact amount”.
- Reality Lab Vol No.011 | “No Germs Detected” — Does That Mean There Were None?
- How to Tell the Presence of a PSLE Science Condition From Its Amount or Level
- How to Tell Measurement Precision From Accuracy in PSLE Science
Vol No.011 asks what a non-detection can and cannot prove. Vol No.094 asks the mirror-image question: after a positive detection, how much numerical detail has really been earned?
Original Reality Lab Case: The Fictional Aqua-Q Test
This case is original and uses fictional substances. It is not medical, health or regulatory advice and does not reproduce a commercial test.
A school science group builds a fictional optical test for Substance Q. The instrument gives a signal score.
| Sample | Signal score | What is known? |
|---|---|---|
| Blank water | 0.8–1.2 | Background signal |
| Low-Q reference | 1.8–2.4 | Q is present, but the amount is near the method’s low end |
| Mid-Q reference | 6.9–7.4 | Q is present and the amount can be estimated reliably |
| Unknown | 2.1 | Signal is above typical blank values |
The unknown’s signal looks more like the low-Q reference than the blank. That may support a detection decision. But suppose repeated low-Q measurements vary enough that the numerical amount is not reliable there. Then “Q detected” can be stronger than “Q = 0.037 units”.
Observed, Claimed and Inferred
| Layer | Statement |
|---|---|
| Observed | The unknown produced a signal score of 2.1 under this method. |
| Compared | The signal is above the usual blank range in this teaching case. |
| Claimed | Substance Q was detected. |
| Stronger inferred claim | The exact amount of Q is known to three decimal places. |
The last step needs its own evidence.
Detection and Quantification Answer Different Questions
- Detection question: Is there enough evidence to distinguish a target-related signal from background under the method?
- Quantification question: Can the method assign a numerical amount with acceptable measurement quality over this region?
The two questions are connected, but they are not identical. A method can sometimes recognise a weak signal before it can measure that weak signal precisely enough for a trustworthy amount.
Why Software Can Make the Difference Hard to See
Modern instruments often calculate a number automatically. That number may be mathematically generated even when the signal lies near a method boundary where uncertainty is large. A dashboard can therefore look more certain than the evidence.
This is similar to Reality Lab Vol No.092, where extra decimal places can appear because calculation software prints them. Vol No.094 adds another question: is the signal even in a region where the method can support reliable numerical measurement?
The Blank Matters
A test does not operate in a world of perfect zero. Instruments, materials and surroundings can produce background signals. That is why scientists compare target-like signals with blanks, controls or appropriate background measurements.
If the unknown is barely different from a variable background, confidence in detection is weaker. If it is clearly separated from background but still near the lower end of reliable numerical measurement, detection may be defensible while exact quantity remains uncertain.
Worked Case 1: A Colour Test Shows a Faint Positive
A fictional strip changes colour when Target R is present. A faint colour change is visible. The package’s validated use might support a positive/negative decision, but the darkness of that colour may not have been calibrated well enough to claim “there are exactly 4.7 units”. Presence and amount must not be fused.
Worked Case 2: A Sensor Gives a Numerical Readout
A sensor displays 0.014 units. Repeated measurements of a low reference vary from 0.006 to 0.022. The printed number looks exact, but the repeat evidence shows that small numerical differences in this region are unstable. A bounded detection statement may be stronger than a fine-grained quantity claim.
Worked Case 3: Strong Signal, Stronger Quantity Evidence
An unknown produces a signal well within the region covered by reliable reference standards. Repeated measurements are close, the calibration is appropriate, and controls behave as expected. Now the numerical amount has much stronger support.
The lesson is not “never trust a number after detection”. It is “earn the number with the right evidence”.
Worked Case 4: Detection in One Sample Does Not Measure the Whole Object
A swab from one corner of a fictional surface gives a positive signal. That result belongs first to the sampled material and method. It does not automatically tell you the total amount across an entire room, object or production batch. Sampling scope and numerical measurement are separate questions.
What Evidence Would Strengthen a Quantity Claim?
- Reference standards that cover the relevant amount range.
- Repeated measurements showing acceptable variation in that range.
- Appropriate blanks and controls.
- A defined measurement procedure and units.
- A calibration relationship that is valid where the unknown lies.
- Sampling that matches the scope of the public claim.
- Uncertainty or method limitations visible when the difference matters.
What Would Weaken It?
- The signal is barely above a variable background.
- The method has only been validated for yes/no detection in that low region.
- The unknown lies outside the calibrated range.
- Software prints many digits without measurement-quality evidence.
- A single small sample is used to claim a total amount for a much larger system.
- No repeat, blank or reference evidence is shown.
Tempting Reasoning That Fails
- “Detected means measured exactly.” Detection and quantification have different evidential demands.
- “A number on a screen must be quantitative evidence.” Software can calculate a number even where the method is not reliable enough for that detail.
- “If exact quantity is uncertain, detection is meaningless.” Not necessarily. A method can support a bounded yes/no conclusion before a precise amount.
- “If it is present, more signal always means proportionally more substance.” That relationship needs calibration and a valid measurement range.
How Far Can the Conclusion Travel?
A careful result may read: “A target-related signal was detected in this tested sample using this method.” A stronger result may add a numerical amount when the method has earned that step.
Scientific strength comes from matching the sentence to the evidence, not from choosing the most impressive wording.
PSLE-Style Transfer Case
A fictional light sensor is used to detect Substance T. Its blank readings are usually between 0.9 and 1.1 units. A sample gives 1.8 units. The method documentation says signals above 1.5 can support detection, but reliable numerical measurement begins only above 3.0.
Question: Which conclusion is better supported?
Reasoned answer: The sample supports a detection of T under this method because its signal is above the stated detection criterion. It does not support a reliable numerical amount using the same method because the signal is below the method’s reliable quantification region.
Explained Practice
Practice A: A test is positive but provides no calibrated numerical range. What can you safely claim? The target-related signal was detected under the method, not an exact amount.
Practice B: A second test gives a strong in-range signal and repeated results agree closely. What becomes stronger? The case for a defensible numerical quantity.
Practice C: A product infographic says “Detected at 0.002 units” but gives no method limits. What should you ask? Whether 0.002 lies in a region where the method can reliably quantify, not merely detect.
Delayed Independent Return: The D-Q Check
- D — Detection: Is the signal distinguishable from background?
- Q — Quantity: Can the method measure the amount reliably here?
- R — Range: Is the unknown inside the supported calibration range?
- S — Sample: What material was actually tested?
- C — Claim: Does the sentence stay within those boundaries?
Parent and Tutor Teaching Guide
Draw a simple traffic-light diagram with three regions: background, detectable-but-not-reliably-quantified, and reliably quantified. Give learners fictional sensor readings and ask them to write the strongest sentence each result earns.
Avoid turning the lesson into memorisation of laboratory terminology. The durable habit is simpler: “Did we find evidence of it?” and “How much is there?” are two questions.
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 — Limit of Detection
- National Institute of Standards and Technology — Quantitation Limit
NIST distinguishes detection limits from quantitation limits: a detectable target and a reliably measured amount are related but different measurement claims. The PSLE Science transfer is to recognise that difference whenever a real-world report jumps from “found” to “how much”.
The Quiet Return
Finding a signal is evidence.
Measuring its amount is another scientific job.
Never let one successful question pretend it answered the next one too.
