PSLE Science Reality Lab Vol No.094 | “Detected” — Does That Tell You How Much Is There?

PSLE-SCI-REALITY-0094

Wait, What? “We Found It” and “We Know How Much” Are Two Different Scientific Claims

Imagine a small laboratory report beside a product comparison. A green tick appears next to one line:

Substance R: DETECTED

Then a caption beneath the result says, “The sample contains a large amount of Substance R.”

The first sentence may be supported while the second is not. A test can sometimes distinguish a signal from background well enough to say that a target was detected, yet still lack enough measurement quality to state its amount with useful accuracy and precision. That difference matters whenever a scientific result moves from a laboratory screen, sensor display or test strip into an infographic, advertisement, headline or comparison table.

This Reality Lab is about one transferable habit: do not let the word “detected” quietly turn into a stronger quantity claim.

Quick Answer

Ask what the test actually reported: presence, absence, a category, or a numerical amount.
Find the background or blank behaviour. A real signal must be distinguishable from ordinary noise or contamination.
Separate the ability to detect a target from the ability to quantify it reliably.
Check whether the reported amount lies inside the method’s useful measurement range.
Keep the conclusion at the strength of the evidence: “detected” is not automatically “a lot”, “dangerous”, “effective”, or “important”.

The Exact Learner Job This Page Owns

This page owns a real-world evidence-transfer problem: a communication object uses a positive detection result as though it automatically provides a reliable amount.

It does not replace the main PSLE Science owners for observation versus inference, signal versus noise, fair testing, measurement range, accuracy, precision, controls, evidence selection or conclusion writing. Those skills remain separate. Reality Lab applies them to a claim that has crossed from a measurement system into public communication.

Original Reality Lab Case: The Pond-Water Sensor

This is an original composite teaching case, not an examination question and not a copied commercial test.

A class uses a fictional optical sensor to check pond-water samples for Marker M. The instrument produces a signal value. When a blank sample containing no added Marker M is tested many times, its readings are usually between 0.0 and 2.0 units because instruments and samples are never perfectly noiseless.

Sample	Sensor signal	Report
Blank A	1.2	background
Blank B	1.6	background
Pond sample P	4.8	Marker M detected
Pond sample Q	42.0	Marker M detected and amount measurable within calibrated range

Both P and Q produce signals that the method regards as detections. But their evidential jobs are not identical. For P, the signal may be far enough above background to support “detected” while still being too close to the lower end of the method for a precise numerical amount. Q may sit comfortably inside a region where the instrument has been shown to convert signal into amount reliably.

A poster that labels both samples simply “positive” can be reasonable. A poster that assigns P an exact concentration without evidence that the method can measure that low reliably may be over-claiming.

Observed, Reported, Claimed and Inferred

Layer	What it could mean
Observed	The instrument produced a signal of 4.8 units for sample P.
Method interpretation	The signal met the method’s rule for “detected”.
Possible supported claim	Marker M was detected in this tested sample using this method.
Stronger inference requiring more evidence	The sample contained exactly 0.037 mg/L of Marker M.
Still stronger unsupported leap	The pond therefore contains a large or harmful amount everywhere.

The scientific habit is to notice each step. The instrument does not speak in headlines. People translate a signal into a result, then translate the result into a claim. Every translation can add assumptions.

Detection Is a Decision About a Signal

Real measurement systems have background variation. A blank may produce a tiny response because of instrument electronics, sample handling, environmental conditions, trace contamination or ordinary measurement noise. For that reason, a scientist does not usually treat every non-zero reading as proof of the target.

A detection rule asks whether the observed response is sufficiently distinguishable from background under the method’s defined conditions. The exact mathematics can become advanced; a Primary 5/6 learner does not need to calculate specialist laboratory limits. The transfer idea is simpler: the test needs evidence that the signal is meaningfully different from ordinary background.

NIST describes a limit of detection in terms of the point below which analytical signal cannot be reliably distinguished from instrument background noise. That is a useful reminder that detection is about recognising a signal against noise, not automatically about measuring the amount with high precision.

Quantification Is a Different Job

To quantify means to assign an amount or concentration numerically. That requires more than deciding that a signal is present. The relationship between signal and amount must be good enough, the method must be operating in a suitable range, and the result must have acceptable accuracy and precision for the intended use.

NIST defines a quantitation limit as the minimum amount that can be quantified with acceptable accuracy and precision. The important learner-level distinction is therefore:

Detection question: Is there enough evidence to distinguish the target’s signal from background?
Quantification question: Is there enough measurement quality to say how much is there?

A test can sometimes answer the first question before it can answer the second well.

The Representation Check: What Did the Graphic Quietly Add?

Scientific communication often compresses a complicated measurement into a simple icon. That can help readers, but compression can hide what kind of claim was actually supported.

A green tick may mean only “detected”.
A long bar may visually suggest “a lot” even when the result was merely qualitative.
A percentage may imply a numerical estimate that the method never established.
A colour scale may suggest precise ordering among samples even when several are simply above a detection threshold.
A sentence such as “high levels found” adds a magnitude judgement that needs a numerical basis and a defined comparison.

Read the legend, units and method note before letting the design of the graphic become the evidence.

The Baseline Check: Compared With What?

The word “detected” is not automatically the same as “high”. High compared with what? A blank? Another sample? A normal environmental level? A product specification? A scientific guideline? These are different comparison jobs.

Reality Lab does not ask Primary students to decide specialist health, legal or regulatory thresholds. It asks them to notice when a communication object has changed the comparison without showing its evidence.

The Method Check: Could the Test Measure This Region Reliably?

Suppose an instrument has been tested thoroughly with standards between 10 and 100 units. A sample produces a very small signal corresponding roughly to 2 units if a line is extended downward. Is “2.000 units” automatically justified? No. The method may be able to flag a weak presence but not quantify that region reliably. The communication should respect the method’s demonstrated capability.

This is where measurement range, calibration, background, repeatability and sample preparation all matter. The public label may contain only one number, but the number inherits the strengths and limits of the whole measurement chain.

Alternative Explanations for a Positive Signal

A positive response can be scientifically useful without being infallible. Depending on the method, possible alternatives can include contamination, another substance producing a similar response, sample carryover, a damaged sensor, an incorrect blank, unusual environmental conditions or a processing error.

That does not mean “never trust tests”. It means tests become trustworthy through controls, validation, suitable reference materials, repeated checks and methods designed for the question being asked.

What Evidence Would Strengthen a Detection Claim?

A clearly defined method and target.
Blank or background measurements showing ordinary signal behaviour.
A positive control or reference showing that the method can respond when the target is present.
Replicate measurements that behave consistently enough for the intended decision.
Evidence that likely interferences were considered.
Sample provenance showing what was actually tested, when and how.

What Evidence Would Strengthen a Quantity Claim?

A demonstrated relationship between instrument response and known amounts.
A sample result that lies inside the method’s useful quantitative range.
Measurement uncertainty or precision information appropriate to the claim.
Suitable calibration and quality-control checks.
Units and sample basis stated clearly.
Enough numerical resolution to support the difference being discussed.

Worked Case 1: “Detected in 8 of 10 Samples”

An infographic says Marker M was detected in eight of ten samples. What can we infer? We can discuss the detection frequency in those ten tested samples, assuming the method and reporting are sound. We cannot automatically infer the average amount, the amount in the two non-detections, or the level in every untested sample from the larger environment.

Worked Case 2: “Detected, Therefore More Than Product B”

Product A produces a positive detection. Product B also produces a positive detection. A social-media post says A contains more because its test strip looks darker. Before accepting that, ask whether strip darkness is validated as a quantitative scale, whether lighting and timing were controlled, and whether the method was intended only to classify positive versus negative. A visual difference is not automatically a calibrated amount difference.

Worked Case 3: “Not Quantifiable” Does Not Mean “Not Present”

A report may say a target was detected but below the level at which the method reports a reliable numerical amount. That is not logically identical to “zero”. It means the evidence supports a weaker claim than a precise concentration.

This is an excellent example of scientific restraint: not every uncertainty must be turned into certainty. Sometimes the most accurate communication is a bounded statement about what the method can and cannot establish.

Worked Case 4: The Single Dramatic Result

One sample gives a much stronger signal than the others. A headline says, “Huge levels discovered.” A careful learner asks whether the sample was retested, whether contamination was ruled out, whether the value lies within the calibrated range, whether nearby samples agree, and whether the sampling method makes this one result representative of anything beyond itself.

Tempting Reasoning That Fails

“It was detected, so there must be a lot.” Detection can establish presence without establishing a large amount.
“The reading is non-zero, so the target is definitely present.” Background and noise must be considered.
“Below the quantitation limit means zero.” It means the method does not support a reliable numerical amount below that region.
“A darker colour always means more.” Only if the method has validated that visual response as a quantitative relationship under the relevant conditions.
“The laboratory used a machine, so the result is exact.” Instruments still have range, calibration, interference and uncertainty limits.

How Far Can the Conclusion Travel?

The safest conclusion follows the evidence object closely. If one method detected Marker M in one sample, the conclusion belongs first to that sample, that method and that time. To travel farther—to the whole batch, the whole pond, a product category, a health implication or a universal rule—the claim needs additional evidence.

PSLE Science reasoning already trains this discipline. Evidence has scope. Variables matter. Observations and inferences are not identical. Conclusions should not outrun the investigation. Reality Lab simply carries those habits into the labels, dashboards and claims learners meet outside worksheets.

PSLE-Style Transfer Case

A fictional sensor is used to test water for Indicator Z. Blank samples usually give readings between 0.2 and 0.8 units. A test sample gives 2.5 units, which the validated method classifies as “detected”. The method only reports reliable numerical amounts for signals from 5 to 50 units.

Question: Which conclusion is better supported: “Indicator Z was detected in the tested sample” or “the sample contains exactly 2.5 units of Indicator Z”?

Reasoned answer: The first conclusion is better supported. The 2.5 value is the sensor response in this teaching case, not automatically the amount of Indicator Z. The method can classify the signal as a detection, but its reliable numerical amount range begins at a stronger response.

Explained Practice

Practice A: A test result reads “target detected; quantity not determined”. What is the strongest supported claim? That the target was detected under the method’s rules, not that a particular amount was measured.

Practice B: A chart labels all positive samples with bars of different heights, but the laboratory only reported positive/negative classifications. What should you question? Where the bar heights came from and whether the original method produced quantitative results.

Practice C: A sample is below the test’s detection threshold. Can you conclude the true amount is exactly zero? No. You can only say the method did not detect the target at or above its capability under those conditions.

Delayed Independent Return: The P-R-E-S-E-N-T Check

P — Positive: What exactly counted as a positive result?
R — Range: Is the value inside the method’s useful measuring range?
E — Evidence: What blanks, controls and references support the result?
S — Signal: Is the response clearly separated from background?
E — Exactness: Did the test truly support a numerical amount?
N — Narrow claim: What is the smallest statement the evidence definitely supports?
T — Travel: What extra evidence would be needed before generalising farther?

Parent and Tutor Teaching Guide

Create three cards: NOT DETECTED, DETECTED, and MEASURED AS 23 mg/L. Ask the learner to order them by what each statement tells us. Then ask what additional evidence is needed to move from one card to the next. The goal is not laboratory jargon. The goal is evidence strength.

A second useful exercise is to show a fictional product graphic containing a green tick, a dark colour patch and the phrase “high amount”. Ask the learner to circle which parts are observations, which are method decisions, and which are communication choices. This makes the evidence-to-claim pathway visible.

Authoritative Sources

The official Singapore sources frame Primary Science as more than recall: learners apply knowledge and scientific inquiry by interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning. This Reality Lab uses that same habit without inventing a special examiner rule or answer template.

The Quiet Return

Finding a signal is useful.

Measuring an amount is useful.

They are not automatically the same scientific achievement.

When a report says “detected”, ask what the method truly knows next: present, measurable, or both?

Small Group Tutorials