PSLE Science Reality Lab Vol No.072 | “Removes 99%” — 99% of What Was There at the Start, and What Is Left?

PSLE-SCI-REALITY-0072

Wait, What? “Removes 99%” can leave 1 thing behind—or 10,000 things behind—depending on what you started with.

A percentage can look complete even when the scientific story is not.

Suppose a filter removes 99% of a harmless test particle from a sample. If the sample starts with 100 particles, about 1 remains. If it starts with 1,000,000 particles, about 10,000 remain.

The percentage is the same. The remaining amount is not.

This Reality Lab is about a deceptively simple real-world phrase: “removes 99%.” Your scientific job is not to be impressed or suspicious. It is to rebuild the measurement that the percentage came from.

Quick Answer

Before accepting a percentage-removal claim, recover the evidence chain:

Starting amount: What quantity was present before treatment?
Final amount: What comparable quantity was present after treatment?
Same basis: Were the before-and-after values measured in the same units, volume, time window and sample basis?
Percentage: Is the “99%” calculated from those comparable values?
Remainder: What 1% is still left?
Conditions: Under what flow, concentration, duration, material or test setup was the result obtained?
Meaning of removal: Was the target destroyed, trapped, transferred, separated or simply no longer detected in the measured stream?

Reality Lab rule: A removal percentage is a relationship between before and after. It is not a standalone description of the final amount.

What This Guide Owns — and What It Routes Back To

This article owns one real-world evidence-transfer job: how to evaluate a public “removes X%” claim by reconstructing the starting quantity, final quantity and remainder before deciding how strong the claim is.

It does not become a general percentages lesson, a water-treatment owner, a filtration owner or a health-safety guide. It applies existing PSLE Science numerical and evidence skills to a real-world communication object.

The new job here is narrower and more physical: percentage removed → starting evidence → final evidence → remainder → claim scope.

The Original Reality Lab Case: The Blue-Particle Filter Test

Imagine a harmless classroom model. A clear container holds water with tiny blue plastic beads that represent particles. A mesh cartridge is used to remove beads from the flowing water. The team counts the beads before and after.

Measurement	Count
Before filter	10,000 beads
After filter	100 beads

The number removed is 9,900 beads.

The fraction removed is 9,900 ÷ 10,000 = 0.99.

So the test supports a 99% removal result for that bead count under those test conditions.

Now compare a second test:

Measurement	Count
Before filter	1,000,000 beads
After filter	10,000 beads

This is also 99% removal. Yet 10,000 beads remain.

The percentage alone does not tell you the final count.

Observed, Claimed and Inferred

Layer	Example
Observed	Before and after measurements under stated test conditions.
Calculated	Percentage removed from the measured starting quantity.
Claimed	“Removes 99%.”
Possible over-inference	“Almost nothing remains,” “everything is gone,” or “the final amount is always safe.”

The first three can be perfectly valid. The problem arrives when the reader silently replaces “99% of the starting amount was removed in this test” with a much stronger sentence about the final state.

The Denominator Is the Starting Point

For a simple removal-efficiency calculation, the starting quantity acts as the denominator. A general form is:

percentage removed = (starting quantity − final quantity) ÷ starting quantity × 100%

U.S. Environmental Protection Agency guidance uses the same evidence structure for removal efficiency: an influent value before treatment and an effluent value after treatment are compared on a consistent basis.

For Primary 5/6 reasoning, the formula matters less than the scientific identity of each term. You must know what the numbers measure before calculating anything.

“99% Removed” and “1% Remains” Are Linked—But Only on the Same Basis

If 99% of the measured starting quantity is removed, then 1% of that starting quantity remains in the measured output—assuming the accounting basis is consistent.

That sounds obvious, but it becomes powerful when a claim hides the remainder.

Starting amount	99% removed	1% remains
100 units	99 units	1 unit
10,000 units	9,900 units	100 units
1,000,000 units	990,000 units	10,000 units

The table does not say whether any of those remaining amounts is important, harmless, acceptable or unacceptable. That would require a separate scientific or safety question. This Reality Lab stays with the evidence structure: the same percentage can correspond to very different final amounts.

Before and After Must Measure the Same Scientific Quantity

A percentage can look mathematically neat while the comparison is scientifically invalid.

For example, imagine the “before” value is a total particle count in 1 litre, but the “after” value is a concentration per 100 millilitres. Those numbers cannot be inserted into the same subtraction until they are converted to a comparable basis.

Likewise, do not compare:

mass before with concentration after;
count in one volume with count in another volume without adjustment;
one-minute collection before with ten-minute collection after;
one type of particle before with a broader category after;
raw signal before with corrected estimate after unless the relationship is explained.

The calculation inherits the quality of the measurements. A correct percentage formula cannot repair mismatched quantities.

Removal Is Not Always Destruction

Another important word hides inside the claim: removed.

In a filter, particles may be trapped in the filter material. In a settling process, material may move into a sediment or sludge. In another process, a substance may be transformed chemically. These are not the same pathway.

EPA treatment guidance explicitly warns that material removed from a water stream may be transferred to another waste stream rather than simply vanishing. The Primary Science lesson is straightforward: “not in the measured output” does not automatically mean “destroyed everywhere.”

This matters when a claim changes verbs halfway through its reasoning:

measured less in the output;
therefore removed from that stream;
therefore destroyed;
therefore no longer exists.

Each arrow needs evidence. Do not let one percentage carry four different claims.

Worked Case 1: Two Filters, Same Percentage, Different Remainder

Filter A is tested with a starting count of 1,000 particles and leaves 10. Filter B is tested with a starting count of 100,000 particles and leaves 1,000.

Both remove 99%.

If a poster displays only “99% removed” for both, the percentages look identical. If the scientific question is about the final number of particles, the hidden starting counts matter enormously.

You cannot say Filter A is “better” merely because fewer particles remain unless the two tests are comparable in starting concentration, volume, flow and other relevant conditions.

Worked Case 2: Same Filter, Different Starting Amounts

The same filter is tested twice. In Test 1, 500 particles enter and 5 leave. In Test 2, 50,000 enter and 500 leave.

Both give 99% removal. This consistency can be useful evidence about proportional performance across those two tested starting levels.

But do not overextend. It does not prove the filter will remain at 99% for every concentration, flow rate, particle size, temperature or usage duration.

Worked Case 3: The Volume Changed

Before treatment, a 1-litre sample contains 10,000 particles. After treatment, a 100-millilitre sample contains 10 particles. A student says, “Only 10 remain, so removal is 99.9%.”

The quantities are not yet on the same volume basis. If the output concentration is uniform, 10 particles per 100 mL corresponds to about 100 particles per litre. Only then can the litre-based comparison be made.

The key skill is not complicated mathematics. It is preserving what one number means.

Worked Case 4: “99% Removed” From One Size Range

An original test counts particles between 10 and 20 micrometres and reports 99% removal. A label shortens this to “removes 99% of particles.”

The shorter wording may broaden the claim beyond the tested range. The result belongs first to the particles, sizes and conditions actually measured.

This is a classic scope problem: a valid narrow result becomes an invalid broad claim when important qualifiers disappear.

The Baseline Can Make a Percentage Look Dramatic

High percentages naturally attract attention. But the scientific importance of a percentage often depends on the baseline.

Suppose a process removes 90% from a starting amount of 10 units. One unit remains. Another process removes 99% from a starting amount of 10,000 units. One hundred units remain.

Which final output is lower? The 90% case.

This does not make 90% “better” in general. It demonstrates why a relative percentage and an absolute final amount answer different questions.

One Dramatic Test Is Not a Universal Removal Rate

Removal efficiency can vary with test conditions. Depending on the system, relevant variables may include:

starting concentration;
flow rate;
contact time;
temperature;
particle or molecule type;
particle size;
filter loading or age;
sample composition;
measurement method.

You do not need to memorise this list as a universal checklist. Ask the PSLE Science question instead: Which variables could affect the measured outcome in this system?

What Evidence Would Strengthen a “Removes 99%” Claim?

The exact target being measured is named.
The starting quantity and final quantity are available.
Before and after use the same scientific basis.
Sampling volumes, times and units are comparable.
More than one test is reported when repeatability matters.
The tested range of starting concentrations or conditions is stated.
The measurement method is suitable for the target.
Results near the method’s detection or measurement limits are treated carefully.
The claim distinguishes removal from transfer or destruction when that distinction matters.
The conclusion is limited to the tested conditions rather than advertised as a universal constant.

What Would Weaken It?

A percentage with no starting quantity.
Before and after values measured on different bases.
One unusually good trial presented as the product’s permanent performance.
Changing sample volume without accounting for the change.
“Removed” quietly becoming “destroyed.”
A broad claim about all particles or substances when only one type or size range was tested.
No information about what remains after the percentage reduction.

How Far Can the Conclusion Travel?

If a well-run test begins with 20,000 measured particles and ends with 200, then 99% were removed from that measured starting quantity under that test’s conditions.

That statement can be strong.

It does not automatically establish that:

no particles remain;
the final amount is suitable for every purpose;
99% will occur under every starting concentration;
the same percentage applies to every particle type;
the removed material was destroyed;
every unit of the product will perform identically forever.

Scientific scepticism is not saying “the percentage is fake.” It is keeping the percentage attached to the experiment that produced it.

PSLE-Style Transfer Case

A student tests a mesh using identical beads. Before filtering, 2,000 beads are counted in the test container. After filtering, 40 beads are counted in the comparable output container.

The student writes: “The mesh removes almost everything because only 2% remains.”

First calculate carefully. If 40 of 2,000 remain, then 1,960 were removed. 1,960 ÷ 2,000 = 98% removed, so 2% remains.

Now evaluate the words. “Almost everything” may be acceptable as an informal description of the proportion in this test, but the scientifically stronger statement is:

Under the stated test conditions, the mesh removed 98% of the counted beads, leaving 40 of the original 2,000 in the measured output.

That sentence preserves both the percentage and the remainder.

Tempting Reasoning That Fails

“99% means zero.” It means 1% of the measured starting quantity remains on the same basis.
“99% is always better than 90%.” Not if the scientific question concerns final amount and the starting baselines differ.
“The final number is smaller, so I can divide by the final number.” Removal percentage is referenced to the starting quantity.
“The before count was per litre and the after count was per 100 mL, but they are both counts.” They are not yet comparable on the same volume basis.
“Removed means destroyed.” Removal can mean transfer, capture or separation from the measured stream.
“A 99% laboratory result applies everywhere.” Performance belongs first to the tested conditions and target.

Practice 1: Recover the Remainder

A test starts with 50,000 particles and reports 99% removal. How many particles correspond to the remaining 1%?

Answer: 500 particles, if the percentage applies to that count on the same basis.

Practice 2: Same Percentage, Different Final Count

Test A starts with 1,000 particles. Test B starts with 100,000. Both remove 99%. Which leaves more particles?

Answer: Test B leaves more: 1,000 particles versus 10 in Test A.

Practice 3: Find the Mismatch

A report compares 200 mg of material before treatment with 5 mg/L after treatment and immediately states a removal percentage. What is wrong?

Answer: The quantities are on different bases: total mass versus concentration. The sample volume and conversion needed to make them comparable are missing.

Practice 4: Removal or Destruction?

A filter traps beads in its cartridge, so few beads appear in the output water. Can the student say the beads were destroyed?

Answer: No. The evidence shows they were removed from the output stream and captured in the cartridge, not destroyed.

Delayed Independent Return: The BEFORE → AFTER → LEFT Audit

When you next see a large removal percentage, write three boxes:

BEFORE — What exactly was there, how much, and on what basis?
AFTER — What exactly was measured after, using what comparable basis?
LEFT — What amount or fraction remains, and what does “removed” physically mean?

Only then return to the headline percentage.

Teaching Guide for Parents and Tutors

Use counters, beads or paper dots rather than a real chemical or health-related product. Start with 100 counters and remove 99. The learner can see one remaining. Then start with 1,000 counters represented symbolically on a grid and leave 10. Finally, use a drawn “one million” case and leave 10,000.

Ask the learner to explain why the percentage stayed constant while the remainder changed.

Next, deliberately change the measurement basis: “100 particles in 100 mL before; 5 particles in 10 mL after.” Ask whether the two counts can be compared immediately. This moves the learner from percentage arithmetic into scientific quantity control.

Finish with the language shift: put 99 counters in a cup labelled “captured by filter.” Ask, “Were they destroyed?” The physical location makes the difference between removal and destruction memorable.

The teaching target is a compact habit: percentage claims must carry their denominator, their remainder and their physical meaning.

Authoritative Sources

The Quiet Return

“99% removed” can be a useful, accurate summary. The scientific mistake is treating the summary as if it contains information that was never supplied.

A percentage needs a starting quantity. A removal needs a before-and-after comparison. A final claim needs the remainder. And the word “removed” needs a physical meaning.

So whenever a product label, infographic, demonstration or report gives you a dramatic removal percentage, do one quiet thing before reacting: reconstruct what was there at the start—and ask what is still there at the end.

Small Group Tutorials