Did you know? A school water filter can turn a cloudy-looking sample into a clear-looking one without telling us whether the water is safe to drink. That is a striking learning moment: our eyes are useful instruments, but they cannot see every dissolved substance or every disease-causing organism. The best investigation is not the one that announces the biggest success. It is the one that knows exactly what its evidence can support.
Clear filtered water is not proof of drinking safety. Filtration means passing a fluid through a material that retains some substances while allowing other material through. A classroom paper-filter demonstration can show removal of some visible suspended particles. It does not, by itself, demonstrate that dissolved contaminants or harmful microorganisms are absent. The immediate rule is simple: describe visible separation as visible separation, never as a drinking-water certification.
The diagnostic question: what did you actually measure?
Show a learner two transparent cups labelled A and B. A looks brown and cloudy. B has been passed through filter paper and looks much clearer. Ask: ‘What can we conclude?’ A careful answer is ‘The appearance changed, and the sample appears to contain fewer visible suspended particles.’ A careless answer is ‘The water is pure and safe.’ The difference is not vocabulary alone. The second claim needs evidence that the classroom observation did not collect.
Follow with a second question: ‘Would an invisible dissolved dye or salt necessarily disappear in the same paper filter?’ No. Ordinary filter paper is not a universal barrier to dissolved substances. Then ask whether seeing no particles proves that there are no microorganisms. It does not. These checks identify whether the learner has confused a proxy measurement such as visible clarity with the separate target property of drinking safety.
What filtration does and what it cannot establish
In the common school demonstration, the filter is a physical barrier. Some solid particles are larger than openings or become trapped in its fibres. Water passes through and carries along substances that are not caught. Filter performance depends on the material, pore characteristics, substances present, maintenance and exact conditions. A result with filter paper does not automatically describe the performance of a specialised drinking-water treatment system.
Visible cloudiness is often called turbidity, but a simple visual comparison is only qualitative; it is not a calibrated turbidity measurement. Clear appearance can coexist with dissolved chemicals. Microorganisms are too small to judge from appearance alone. Conversely, some visible colour or harmless material does not establish that a sample is dangerous. The right lesson is to test the property we wish to claim.
Worked example: a fictional classroom filter investigation
All samples, quantities and observations in this worked example are invented for teaching. They are not experimental findings, and none of the samples is intended for drinking. Students prepare a demonstration by mixing a little soil into 200 mL of ordinary water in a school science setting. They pass it through a paper filter into a labelled collection beaker. The recovered filtrate measures 175 mL. It looks clearer than the starting mixture. No dissolved-substance test, microbial test, safety certification or calibrated turbidity instrument is used.
Step one: calculate only what the collected numbers support
The measured liquid recovery fraction is 175 mL divided by 200 mL = 0.875, or 87.5%. The unrecovered difference is 25 mL. The liquid may remain in filter paper, equipment or the residue, and handling losses are possible. We cannot infer that exactly 25 mL evaporated, nor that 12.5% of contaminants were removed. A volume difference measures volume recovery, not contaminant-removal efficiency.
Step two: state the observation without expanding the claim
A defensible results sentence is: ‘After paper filtration, 175 mL of liquid was collected from an initial 200 mL sample, and the filtrate appeared clearer in the same viewing conditions.’ The sentence specifies quantity, method and observation. The clarity comparison remains qualitative, and no drinking claim is made. If the starting and ending cups have different shapes or lighting, even the appearance comparison is less reliable; matching containers and backgrounds would improve it.
Step three: name what remains unknown
- Concentrations of dissolved substances were not measured.
- Presence or absence of harmful microorganisms was not tested.
- Neither a pathogen-removal performance standard nor a drinking-water standard was assessed.
- The 25 mL volume difference has several possible causes and does not identify any one cause.
- The class has not measured the filter’s performance across repeated independent trials.
The correct conclusion is not disappointing. It is precise: ‘In this demonstration, paper filtration was associated with visually clearer liquid and 87.5% measured liquid recovery. Safety for consumption was not established.’ A learner who can say that understands both the science and the limits of evidence.
Three questions that must remain separate
- Did the water look clearer? Answer with matched visual observations or a suitable turbidity measurement.
- Did a specified contaminant decrease? Answer only using a test suited to that contaminant, before-and-after comparisons and appropriate controls.
- Is the water safe to drink? Answer using applicable water-quality requirements, validated treatment and testing, and guidance from the responsible water authority. A classroom filter demonstration cannot supply this by itself.
The US Centers for Disease Control and Prevention explains that different water filters address different substances and that appearance, taste and smell are not reliable safety indicators. Its emergency water-safety guidance also explains why filters do not necessarily remove all germ types and why boiling does not remove hazardous chemicals. Emergency treatment instructions vary by contaminant and local advice; do not adapt a classroom demonstration into a home treatment recipe.
Progressive practice: guide, compare and independently verify
Guided practice: circle the overclaim
Give three sample conclusions: (A) ‘The liquid looked clearer after filtration.’ (B) ‘The filter removed every dissolved substance.’ (C) ‘The filtered sample was safe to drink.’ Ask which statement matches a visual-only experiment. A is supported if viewing conditions were comparable; B and C are not. Then ask the pupil what extra evidence each unsupported statement would need.
Paired practice: decide between two explanations
Student One proposes, ‘We lost 25 mL, so the filter removed 25 mL of dirt.’ Student Two says, ‘The collected liquid volume was lower, but we did not measure the amount of trapped dirt or retained liquid separately.’ Ask partners to identify the better explanation and explain why. Student Two wins the evidence test. A mass of dry residue, a volume of liquid and a particle count are different kinds of measurements.
Independent practice: design a better comparison
Imagine two paper-filter designs tested with equally prepared mock muddy-water samples. Sketch a results table with starting volume (mL), collected volume (mL), identical lighting and container details, visible clarity observation, filtration time (s), and a note explaining what was not tested. State one control variable: for example, use the same starting mixture volume and filter area for each design. Plan repetitions and avoid treating three measurements from one single pour as three independent tests.
Common misconceptions and the repairs that work
- ‘Clear means pure.’ Replace with ‘clarity is one observed property, not a complete chemical or microbiological analysis.’
- ‘All filters remove the same contaminants.’ Replace with ‘performance depends on the particular filter and target substance.’
- ‘A larger liquid loss means better purification.’ Replace with ‘measure contaminant removal separately from liquid recovery.’
- ‘No smell means no problem.’ Replace with ‘sensory clues cannot establish all safety properties.’
- ‘The class result proves a household filter works.’ Replace with ‘a classroom apparatus and a drinking-water device have different purposes and validation requirements.’
What parents and teachers can do immediately
Teachers: label all demonstration samples ‘not for drinking’, use a safe teaching setup, require students to separate observations from claims and provide a results-sheet field titled ‘Not measured’. Invite learners to specify a test before allowing claims about any substance. Students should never drink classroom samples. Parents: ask, ‘Which property did you see, and which would need a laboratory test?’ That question is more educational than asking whether the experiment ‘worked’. For an actual drinking-water concern, use the relevant official water supplier or health authority rather than any school experiment.
Observable progress is a pupil who can identify the measurement, calculate recovery with units, flag untested properties, propose a suitable next test without performing unsafe tests, and revise an exaggerated conclusion. This is independent checking: the learner is no longer relying on whether a result looks persuasive.
Limitations: where the classroom model must stop
Paper filtration can be a valuable representation of separation, but it cannot reproduce the full chain of professional water treatment and verification. A household or commercial device may have certified capabilities for particular contaminants; those claims must be checked against the device’s actual certification, condition and operating instructions. Even validated treatment for germs is not a universal remedy for chemical contamination. We should not infer actual drinking safety from classroom appearance or from the fictional numbers above.
Frequently asked questions
If the filtered sample started with tap water, may students drink it?
No classroom sample should be consumed as part of this activity. Classroom handling, soil additions and equipment introduce uncontrolled factors. Use ordinary drinking water separately according to school practices, not the experiment’s beakers.
Does filtration always remove bacteria and viruses?
No. Removal depends on the specific device, pore characteristics, validation and organisms involved. The CDC cautions that many ordinary filters do not remove all types of germs. A generic classroom paper filter is not evidence of disinfection.
Can we call a clearer sample a successful experiment?
Yes, provided success is defined narrowly and in advance: a visible decrease in cloudiness under comparable conditions, with clear acknowledgement of limitations. It is not success at proving drinking safety.
The larger educational aim: conclusions sized to evidence
This skill transfers to claims about food, air, medicines, cleaning devices, science videos and advertisements. We need to ask what a demonstration can actually test. For connected learning, explore how water filtration works and what it cannot remove, and how to detect an instrument’s zero error before interpreting data. The eduKate Education Hub connects these evidence habits across subjects. Our immutable Clementi small-group learning reference shows why diagnosis, carefully sequenced explanation and independent checks matter.
A good young scientist is not embarrassed to say ‘we have not tested that yet’. That sentence is often the beginning of a much better investigation.